Miranda Today
The nature of our project changes in this chapter. The earlier chapters described how and why interrogation law developed the way it did. In chapter 8, for example, we offered theories that, we hope, explain the current state of the Miranda doctrine better than the Supreme Court’s on-again, off-again characterization of Miranda as prophylactic. But in the final two chapters, we will both describe the current law governing interrogation and offer a normative vision of how it can be improved.
We also narrow our scope in this chapter to interrogation problems connected to Miranda. A theme of this book is that something akin to Miranda has been the default position in Anglo-American law since Blackstone’s eighth edition in 1778. The Hawkins-Leach dictum that appeared in 1787 neatly captures Miranda’s key premise—that official interrogation is inherently coercive. For more than a century, English and American courts applied a version of this dictum. Moreover, we discovered an 1828 New York statutory approach, probably influenced by John Graham’s 1823 monograph, that required warnings remarkably like Miranda would require in 1966.
In chapter 8 we focused on Miranda; the Court’s big gun in the attempt to cleanse compulsion from the police interrogation room. To keep the narrative relatively clean and relatively short, we keep the focus on Miranda in this chapter. Miranda was a once-in-a-century attempt to solve the problem of compelled confessions and deserves its own treatment.
As we sought to show in chapter 8, Miranda changed the interrogation world in ways that were dramatic and profound, but ultimately superficial. Imagine telling Wigmore that the Court would require police to warn suspects they had a right not to answer questions and a right to counsel, and then require police to respect their decision to remain silent or to invoke counsel. We also showed in chapter 8 that subsequent Courts rendered the Miranda regime less potent by making the critical moment that of waiver. Warnings make waiver almost always valid, and suspects waive their rights in roughly 80% of interrogations.
Even if one accepts the Court’s modification of Miranda into a weak-force waiver doctrine, it might still fail to achieve the goal of reducing the level of inherent compulsion at the moment of waiver. There are several ways in which Miranda might fail as a weak-force waiver doctrine in routine cases.
First, the warnings themselves might not be comprehensible. One of the best studies on comprehensibility was published by Morgan Cloud and co-authors in 2002.1 The researchers compared subjects who were mentally retarded or borderline mentally retarded—the average IQ was 55.52—with subjects who suffered no mental retardation. The researchers found that the nondisabled subjects exhibited a high degree of comprehension of the rights entailed by the warnings, ranging from 83% to 90% using four broad measures.3 To be sure, this finding likely overstates the comprehension of the typical nondisabled suspect. Suspects under arrest and facing police interrogation are surely more stressed and would be less likely to think clearly than participants in a lab experiment. But the Cloud study nonetheless suggests that not much of the 80% waiver rate can be attributed to incomprehensibility of the warnings. Intuitively, being told that “you have a right to remain silent” seems easily comprehended even if the suspect is under intense stress.
Beyond lack of comprehension, warnings might fail in other ways. Police sometimes undermine the warnings. Charles Weisselberg, Barry Feld, Welsh White, and Richard Leo have, in different venues, made a powerful argument that police often use “softening up” techniques that drain the warnings of their effectiveness.4 These techniques include suggesting the value of talking to police and minimizing the importance of warnings by blending them with booking questions or a nonthreatening conversation. Perhaps the most effective strategy is to tell the suspect that he will not be able to tell his side of the story until the bureaucratic ritual is satisfied.
Based on the research and observations of others, and his own examination of interrogation-training materials in California, Weisselberg sums up the evidence as follows:
We do not today have a clean separation between administration of Miranda warnings and the use of interrogation tactics, at least not in the way the Miranda Court envisioned. Observational studies and my review of training materials provide significant evidence that the warnings and waiver regime has moved at least partway into the interrogation process, contrary to the “time out” from the pressures of interrogation the Court imagined. Officers may use pre-Miranda conversation to build rapport, which is important to obtaining a Miranda waiver and—eventually—a statement. Officers may also downplay the significance of the warning or portray it as a bureaucratic step to be satisfied before a conversation may occur. There is also evidence that police often describe some of the evidence against suspects before seeking waivers. A few cases have approved extreme versions of this tactic.5
And there are two doctrinal wrinkles that potentially drain the warnings of their ability to negate the inherent compulsion of police interrogation. One is whether courts police the waiver and invocation “boundaries” to keep those moments from becoming tainted by compulsion. Though the law on this point was in flux until 2010, Berghuis v. Thompkins made clear that the Supreme Court does not expect much in terms of boundary patrolling.6 Indeed, in removing the last residue of the “heavy burden” waiver language from Miranda doctrine, Thompkins is perhaps the most significant Miranda case yet decided. As we noted in the last chapter, police carefully read Thompkins his rights, which included “the right to decide at any time before or during questioning to use your right to remain silent and your right to talk with a lawyer while you are being questioned.”7 After almost three hours of questioning during which Thompkins said nearly nothing, the interrogator found his weak spot. He asked if he believed in God and if he prayed to God.8 When Thompkins answered yes to both, tears welled up in his eyes, and the officer asked: “Do you pray to God to forgive you for shooting that boy down?” Thompkins answered, “Yes.”
Thompkins made two arguments. One was that his lengthy silence had somehow invoked his right to remain silent. The Court rejected that argument with relative ease—though the dissent found it “counterintuitive[]” that a suspect had to talk to invoke the right to remain silent.9 But if silence can invoke Miranda, how would courts draw the line when silence matured into an invocation? Three hours? Two? One? Would it matter if, as Thompkins did, the suspect responded to two or three questions that did not relate to the investigation? What if the questions were about the investigation?
That Thompkins lost on this issue is thus not surprising. The statement of the rule was, however, significant. “The Court has not yet stated whether an invocation of the right to remain silent can be ambiguous or equivocal, but there is no principled reason to adopt different standards for determining when an accused has invoked the Miranda right to remain silent and the Miranda right to counsel at issue in Davis.”10 An unambiguous invocation standard is easier for courts to apply because it “results in an objective inquiry that ‘avoid[s] difficulties of proof and … provide[s] guidance to officers’ on how to proceed in the face of ambiguity.”11
But we are left with the nagging doubt that substantial compulsion will often attend interrogations in which police ignore attempted invocations of the right to silence that are less than crystal clear. In United States v. Banks, for example, a detective read the suspect his Miranda rights in the back of the squad car and asked him if he understood them.12 He said that he did. The officer then read a waiver form and asked if he was willing to waive his rights. He responded: “Get the f—out of my face. I don’t have nothing to say. I refuse to sign [the waiver form].”13 The federal magistrate held that this was not a clear assertion of his right to remain silent but was, instead, “the statement of somebody who is angry about what is happening to them, which makes a certain amount of sense under these circumstances.”14 Why it was not both an angry response and an invocation is not clear to us, but the Seventh Circuit affirmed. Presumably, Thompkins validates the outcome in Banks. Neither “I don’t have nothing to say” nor “I refuse to sign” the waiver form is an explicit invocation of the right to remain silent, though writing that makes us wonder what happened to the spirit of Miranda.
On the way to the federal building, the detective told the suspect that his co-conspirators referred to him behind his back as “Big Dumbo.”15 At the federal building, he was given Miranda warnings again but was not presented a waiver form or asked whether he wished to waive his rights. During the subsequent interrogation, police played a tape of Banks’s co-conspirators making fun of him, and he made incriminating statements. Was this waiver? That question returns us to Thompkins.
Thompkins’s second Miranda argument was that the State had failed to sustain its burden of demonstrating waiver. Implicit in this argument was that, at a minimum, the State must show an express waiver that is distinct from the act of answering questions. This argument follows from Miranda itself: “An express statement that the individual is willing to make a statement and does not want an attorney followed closely by a statement could constitute a waiver. But a valid waiver will not be presumed simply from the silence of the accused after warnings are given or simply from the fact that a confession was in fact eventually obtained.”16 The Thompkins Court acknowledged: “Some language in Miranda could be read to indicate that waivers are difficult to establish absent an explicit written waiver or a formal, express oral statement”17 and then proceeded to rewrite that part of Miranda. The rule now is: “Where the prosecution shows that a Miranda warning was given and that it was understood by the accused, an accused’s uncoerced statement establishes an implied waiver of the right to remain silent.”18
Return to Banks. Because the suspect admitted that he understood his Miranda rights and because his statement appeared uncoerced, the suspect waived Miranda even though he made no express waiver. His earlier statements about having nothing to say probably fail the Thompkins “unambiguous” test for invocation. Thompkins changed, perhaps clarified, a lot about how Miranda is understood.
Does a Miranda world with Thompkins boundaries permit too much compulsion at the critical points of waiver and putative invocation? That, of course, is a normative question. But imagine what might have been going through the mind of the suspect in Banks. He had told the officer, about as clearly as possible, to leave him alone, that he had nothing to say, and that he was not going to sign a waiver. He is then told by the detective that his co-conspirators are making fun of him. When next given warnings, he is not asked to sign a waiver. Nor does he give an express waiver before the detective plays a tape of his co-conspirators making fun of him and then begins to interrogate him. The suspect might have changed his mind about answering police questions. Or he might have resigned himself to having to submit to police questioning.
It seems inevitable that some suspects will react with resignation to a detective’s refusal to take seriously the “I won’t talk to you” signal. This was the point that Justice Souter made in his separate opinion in Davis—where the Court required an unambiguous request for counsel before it counted as invocation. In an opinion joined by three other justices, Souter wrote:
When a suspect understands his (expressed) wishes to have been ignored (and by hypothesis, he has said something that an objective listener could “reasonably,” although not necessarily, take to be a request), in contravention of the “rights” just read to him by his interrogator, he may well see further objection as futile and confession (true or not) as the only way to end his interrogation.19
It thus seems to us that the Thompkins boundaries will not always suffice to negate the inherent compulsion of police interrogation at the waiver point. The number of cases of boundary failure is yet unknown. One effect of Thompkins is to avoid what would otherwise be a rich irony. If courts were left to puzzle whether to credit ambiguous requests for counsel or to remain silent, they would be back in the game of trying to infer the suspect’s will. By requiring unambiguous signals, the Court has greatly reduced the “mind-reading” required of courts, at a unquantified cost in compulsion.
And there is another doctrine that permits police to avoid the serious waiver moment that the Miranda Court contemplated, even on a weak-force understanding of Miranda. The narrow statement of the holding in Miranda is limited to custodial interrogation. Thus, if courts conclude that what the police did is not interrogation, or that the suspect was not in custody, then there is no need for even a softened-up waiver. In Oregon v. Mathiason, for example, the Court held that a parolee was not in custody even though he was questioned in a police interrogation room and had been told, falsely, that his fingerprints were found at the scene of the crime.20 The key to the per curiam opinion was that the police told him he was not under arrest and, in fact, allowed him to leave when the interrogation was over.21
As Justice Marshall pointed out in his dissent, however, the “coercive elements” acting on Mathiason were “pervasive” even if he was not technically in custody.22 Marshall pointed the way to an evolving Miranda doctrine that would count coercive pressure without regard to custody, based on substance over form: “[F]aithfulness to Miranda” required the Court to consider whether the particular police conduct had the “coercive aspects” of custodial interrogation rather than creating bright-line categories that would be exempted from Miranda.23 But Marshall’s argument was based on the strong-force understanding of Miranda, which by the time of Mathiason in 1977 no longer commanded a majority.
The custody-interrogation boundary thus permits courts to ignore what would otherwise be an invocation. In Burket v. Angelone, the murder suspect agreed to come to the station “to help with the investigation.”24 During the “interview,” police lied and told him that a hair sample matching his had been found in the victims’ residence.25 He then admitted being in their house on the day they were killed and, two minutes later, said “I’m gonna need a lawyer.” Police responded by telling him that he was not under arrest and was free to leave at any time. “Burket continued to talk with the detectives” and incriminated himself. The Fourth Circuit held that because he was not in custody, he had no Fifth Amendment right to counsel.26 Despite a clear statement that he wanted counsel, Burket lost his motion to suppress, his subsequent appeals, and his petition for a writ of habeas corpus.27
George Thomas analyzed 211 Miranda cases he drew at random from West-law in 2002.28 There is a selection bias in choosing cases from Westlaw because almost all—94%—of the cases were appellate cases.29 But the bias should be in the direction of finding more claims of Miranda violations than actually occur because most cases without plausible arguments will not reach trial or, if they do, will not be appealed. In the Westlaw data set, no warnings were given in 60 of the 211 cases, but statements were suppressed in only 10 of them.30 Thus, in over 80% of the no-warning cases, lower courts read Supreme Court doctrine as permitting police not to give warnings.
Another artifact of the weak-force understanding of Miranda appears in Thomas’s Westlaw survey in the way the courts described the involuntariness claims—both of the waiver and of the subsequent statement. The claims in almost all cases lacked details that might prove involuntariness and had an almost pro forma quality to them. “The typical claim was that the waiver or confession was involuntary or that the police used trickery or deception, with few underlying details as to what constituted the coercion or the trickery.”31
Reading these cases, one gets the impression that the lawyers often made an involuntariness claim as a “Hail Mary” pass or to protect against a claim of ineffective assistance. This reading is borne out by an American Bar Association report, “Criminal Justice in Crisis,” published in 1986.32 Defense lawyers surveyed reported that “[v]ery rarely” did they win a Miranda suppression motion. One assistant public defender said that he “[s]eldom, if ever” had a motion to suppress granted. “The difference between zero and none, lately. … And I don’t really have the complaint, today, about Miranda not being given as often, near as often as I did years ago. But I don’t see the courts throwing out Miranda cases.” It thus appears that defense lawyers perceive the chance of winning a suppression motion as nil when police give the warnings. This could either be because police are at least formally playing by Miranda’s rules or because lawyers are cynical about the willingness of courts to exclude confessions—or both, of course.
As Thompkins makes plain, the evolution of Miranda has happened on the watch of Courts that construed Miranda narrowly even as a waiver or notice doctrine. Thus, if Thomas is right that Miranda is about giving notice, or if Dripps and Rosenthal are right that Miranda represents merely the standard application of waiver law, the reader would be entitled to conclude that Miranda is less than a robust success. One could not, however, conclude that Miranda has been a complete failure. In Thomas’s 2002 study, suspects invoked Miranda 14% of the time and did not challenge the validity of the waiver in another 40% of the cases.33
One could conclude, using these data, that Miranda “worked” half the time. Invocation is an act that demonstrates that the inherent compulsion of the police-dominated atmosphere has been ameliorated. A suspect who tells police he will not answer questions or wants a lawyer is directly challenging the police authority to conduct the interrogation as they please. An unchallenged waiver should also count as a success for the weak-force Miranda theory. If the defendant felt he was tricked or compelled, presumably he would have challenged the State’s claim that he waived his rights.
Whatever the failures of Miranda as ensuring notice or a valid waiver, failure is magnified when the suspect lacks the full capacity to process information. Though lack of capacity exists in other contexts (intoxication, lack of language facility), the most vexing contexts are when suspects are young or suffer from mental retardation. We turn now to those contexts.
Part of our argument in this book is that the Court inadvertently lost most of the value in the voluntariness test that Miranda sought to supplement. One largely lost value is the due process focus on the characteristics of the suspect, which is almost always muted or nonexistent in Miranda cases. As noted in the last section, Morgan Cloud and co-authors conducted an empirical study of mentally retarded subjects who were asked the meaning of the Miranda warnings. Only 31% of the subjects indicated an understanding of the right to counsel, which was the highest level of rights’ comprehension.34 To take another example from the Cloud study, the subjects were asked whether two sentences—“You do not have to make a statement and have the right to remain silent,” and “You do not have to say anything about what you did”—have the same meaning.
[T]he control group gave the correct answer 91 percent of the time. In stark contrast, the retarded subjects were correct only 10 percent of the time. Similarly, only 20 percent of the disabled subjects answered correctly the question, “Do you have to tell the police what happened?” compared to 96 percent of the controls. Random guessing would have yielded 50 percent correct. Only 31 percent of the disabled gave correct answers when asked, “Do you have to answer the police if they ask you any questions? “[—]far lower than the 50 percent from random guessing and the 82 percent correct score obtained by the control group.35
The authors concluded that their empirical research showed that “retarded people simply do not understand their Miranda rights. They do not understand the words comprising the warnings. They do not understand the rights themselves. They do not understand the legal context in which the rights arise. Miranda fails to protect the rights of mentally retarded people.”36 Moreover, the study’s findings “suggest that people who are not classified as retarded, but who have low IQs, also may not understand the warnings.” The study continued:
Both our theoretical and empirical analyses lead to the conclusion that the language of the relevant constitutional doctrines is meaningless as it has been employed with mentally retarded suspects. The words of the Miranda warnings themselves are “meaningless” to mentally retarded suspects, who simply do not understand them. The words used to define the waiver test are “meaningless” in a different sense: they are applied to mentally retarded suspects in ways that contradict the very meaning of the words by which the standard is defined. Disabled suspects’ waivers of the rights described in the Miranda warnings are “voluntary, knowing, and intelligent” only if we are willing to manipulate and distort the very meaning of these terms. Finally, the words of these confessions may themselves be meaningless. The cognitive and social weaknesses experienced by mentally retarded people ensure that some of their confessions will be false. When subjected to the pressures of custodial interrogation, mentally retarded people are more likely than others to confess to crimes they did not commit.37
In Miller v. State, the trial judge found as a matter of state law that Miller was “a mentally retarded individual”38 and thus not eligible for the death penalty. Yet the trial judge and the Indiana Supreme Court not only found his Miranda waiver voluntary but also voluntary beyond a reasonable doubt—the standard in Indiana.39 The court fell into a familiar trap when it noted in connection with the waiver analysis: “There is no allegation or indication that police knew that he was mentally retarded.”40 What difference could that possibly make in whether the suspect understood the rights he was “waiving”? Either he understood the warnings well enough to waive his rights or he did not. The officer’s misconduct is relevant in civil rights claims for damages but not, we think, on the waiver or voluntariness questions.
One might have expected the state courts in Miller to inquire into the level of mental disability and discuss whether a suspect who functioned at that level could validly waive Miranda. The trial judge found that Miller’s IQ was “approximately 70 to 75 or below,”41 and 70 is typically used as the borderline measure of mental retardation.42 As we will see shortly, the United States Supreme Court under the old due process test took pains to account for a suspect’s mental limitations. What makes Miller even odder is that the interrogation does not appear to comport with Miranda, at least the pre-Thompkins version, even if the suspect had not been retarded. When Miller heard that his name was mentioned on television in connection with the rape and murder of a seventy-one-year-old woman, Miller “voluntarily went to the police station to ‘get it cleared up.’”43 He arrived at 5:30 P.M. and was placed in an interview room that locked automatically from the outside. He waited there, alone, for two hours.
A detective joined Miller in the locked interrogation room and began to gather “background and preliminary information.”44 At some point during this questioning, the detective “considered the defendant to be a suspect and orally informed the defendant of his rights.” The court mentions no evidence of waiver. During the next six hours, the detective’s “questioning became more focused and included confronting the defendant” with assertions that he was guilty. The detective showed Miller a police report that falsely characterized the death as accidental. He also lied about the evidence against Miller. He told Miller that witnesses had seen him in the hallway outside the victim’s office, which was untrue. He showed Miller a fingerprint card and computer printout that proved Miller’s fingerprints were found in the victim’s office; no fingerprint match had yet been made. Shortly before 1:00 A.M., after six hours of intense questioning, Miller admitted involvement in the victim’s death.
At that point, the detective left Miller alone for forty-five minutes.45 When he returned, he told Miller that he was under arrest and that he wanted to put the rest of the interrogation on tape. At this point, after six hours of interrogation, after Miller admitted being in the victim’s office when she fell and died, the detective went over the Miranda warnings thoroughly, pausing after each right to ask him if he understood the right.46 As to each right, Miller acknowledged that he understood. In response to the detective’s concluding question: “What does it mean to you when I tell you your rights?” he responded, “It means that if I didn’t want to, you know, say anything, that I can talk to an attorney or I could, you know, come on with (inaudible) you know, to get this cleared up.”
This is a powerful piece of evidence that, at the moment of the second set of warnings, Miller more or less understood his rights. Moreover, Miller invoked his right to counsel after making incriminating statements in the second phase of the interrogation.47 The state supreme court seemed to assume that invocation entailed understanding not only the right to counsel at that moment but also the right to counsel and to remain silent for the six hours the police relentlessly questioned him before the second set of warnings. But this is, to say the least, a contestable proposition.
To compare Miller with pre-Miranda cases is both instructive and sad. In Fikes v. Alabama, the Court suppressed a confession made by a black man who was “uneducated” and “of low mentality, if not mentally ill.”48 The only coercive factor was interrogation over several days in isolation.49 Miller was interrogated for over six hours in a locked interrogation room. Unlike Fikes, there was no need to speculate about whether Miller was mentally incapacitated. He was legally determined to be mentally retarded. We will return to Fikes later in the chapter, when we show the desirable flexibility of the due process test that the Court relied on before it decided Miranda. For now, it suffices to say that Miller would have won a due process claim if decided by the 1957 Supreme Court.
Indeed, when the Court was presented evidence in 1960 that a suspect had been diagnosed as mentally ill, it unanimously reversed his conviction even though he was questioned for only five or six hours.50 The striking aspect of Blackburn v. Alabama is that almost all of the opinion is about the suspect’s mental capacity; very little is said about the interrogation. No one disputed the sheriff’s claim that Blackburn was not “threatened . … in any way.” Nor was there evidence that the police lied to him about evidence that “proved” his guilt. Miller’s problem, it seems, was being judged by Miranda 2002 standards rather than the Supreme Court’s 1950s due process standards.
A similar capacity issue arises with juveniles. Here, too, the Court has also largely abandoned the protections of the pre-Miranda due process cases. In Haley v. Ohio, five members of the 1948 Court said that advising a fifteen-year-old of his right to remain silent was not enough to make his confession voluntary.51 The plurality rejected the proposition that “a boy of fifteen, without aid of counsel, would have a full appreciation of that advice” and would know that he had “a freedom of choice.” Justice Frankfurter’s separate opinion stressed that the juvenile had no “appreciation of his rights” and that “the means of asserting them were effectively withheld from him by the police.”52 To be sure, the interrogation was five hours in length and began at midnight,53 so what is remarkable is not the finding of involuntariness but that the plurality did not “give any weight to recitals which merely formalize constitutional requirements.”54
Gallegos v. Colorado was decided four years before Miranda reordered the interrogation universe.55 Over a spirited dissent, the Court held that the use of a confession given by a suspect who was fourteen years old violated due process, even though there was “no evidence of prolonged questioning.”56 Indeed, “the basic ingredients of the confession came tumbling out as soon as he was arrested.”57 While the Court cited instances of police misconduct,58 the most significant factor was Haley’s age.
[A] 14-year-old boy, no matter how sophisticated, is unlikely to have any conception of what will confront him when he is made accessible only to the police. That is to say, we deal with a person who is not equal to the police in knowledge and understanding of the consequences of the questions and answers being recorded and who is unable to know how to protest his own interests or how to get the benefits of his constitutional rights.59
And then, in the next paragraph: “A lawyer or an adult relative or friend could have given the petitioner the protection which his own immaturity could not. Adult advice would have put him on a less unequal footing with his interrogators.”
The concern with the special vulnerabilities of juveniles continued in the 1967 seminal decision of In re Gault, where the Court said: “With respect to juveniles, both common observation and expert opinion emphasize that the ‘distrust of confessions made in certain situations’ to which Dean Wigmore referred. … is imperative in the case of children from an early age through adolescence.”60 The special concern present in Haley, Gallegos, and Gault would end in 1979 in a 5–4 Miranda decision, Fare v. Michael C.61 Sixteen years old, Michael C. responded to the Miranda warnings by asking to speak to his probation officer.62 The police refused the request but repeated that he did not have to talk to them without an attorney present unless he wanted to.63 He agreed to talk.
The California Supreme Court held that the request to talk to the probation officer was an indication that Michael C. did not want to talk to police without help.64 The state court had written in an earlier case that it was “fatuous” to assume that a minor who faces custodial interrogation “will be in a position to call an attorney for assistance.”65 Thus, courts should recognize that the “call for help from the only person to whom he normally looks” for help—his parents in the earlier case—should be taken as an invocation of the right to remain silent. The Michael C. state court extended that holding to a request to talk to a probation officer.66
The Supreme Court reversed, insisting that the only invocations that would produce assistance or require the police to stop questioning were the ones set out in Miranda—a request for counsel or an assertion of the right to remain silent.67 Michael C. did neither and thus there was no invocation. One can quibble with the Court on this point, preferring the state court’s more common-sense view, but the real failure of Michael C. is what happened next. To say that the juvenile did not invoke Miranda is not to say that he waived Miranda. Despite Haley’s admonition that formulaic warnings will not assure a voluntary confession when the suspect is fifteen, Michael C. treated a sixteen year old as an adult without even acknowledging Haley. How did the Court know that Michael C. had waived his Miranda rights? Because there was “no indication that he was of insufficient intelligence to understand the rights he was waiving, or what the consequences of that waiver would be. He was not worn down by improper interrogation tactics or lengthy questioning or by trickery or deceit.”68 But this, of course, is the same standard applied to adults.
We agree with Kenneth King that the standard used to evaluate whether a juvenile has waived Miranda has been “taken directly from adult Miranda jurisprudence and incorporated without modification.”69 Courts thus hold “that a juvenile’s individual character or maturity is irrelevant to the voluntariness analysis and that the use of psychological ploys, which play on a child’s vulnerability and susceptibility to pressure, are unobjectionable.” In sum, “courts have forgotten that children are different from adults and require extra protection.” How could the Court in Michael C. believe that a sixteen year old, who has been denied a chance to speak with his probation officer, possessed an adult’s ability to refuse to talk to police officers? If police are authority figures to adults, imagine how they appeared to Michael C. after they had denied his request for help.
To be sure, the attitude toward juvenile crime has changed dramatically in the last four decades. In re Gault was decided in 1967, just as the attitude had begun to shift from juvenile offenders as children who needed the help of adults to juvenile offenders as dangerous and largely beyond help. These attitudes are necessarily reflected in judicial approaches to the law of interrogation, precisely as our deviance thesis predicts. But some applications of adult standards seem bizarre to us.
Incredibly, the application of adult waiver standards to juveniles has extended to those as young as eleven. We offer three examples. Suspecting that Christopher W. had set a fire at a church, a police officer took him into custody and “after reciting Miranda rights, interrogated him.”70 After about an hour of unsuccessful interrogation, the officer produced a New Testament and asked Christopher if he believed in God, and if he knew that church was God’s home. When the eleven-year-old answered yes, the officer asked him if he could place his hand on the Bible, look the officer in the eye, and say he had not been in the church that burned. At that point, Christopher confessed.
The opinion is a page and one-half, with no dissent and no mention of waiver. When the suspect is eleven years old, a court perhaps should hold that waiver is impossible as a matter of law. If waiver is theoretically possible, it should be the critical issue. The South Carolina Court of Appeals seemed to assume that merely answering questions was sufficient to show that an eleven year old understood and waived his Miranda rights. The state court contented itself with noting that “his testimony shows he is intelligent, quick, and articulate. Furthermore, Christopher, who has had several experiences with the police, testified he understood his right to remain silent and the meaning of his statement.”71 The court’s facile assumption, that an intelligent eleven year old is the best judge of whether he understands his right to remain silent and whether he is able to act on that knowledge, is stunning.
In Matthews v. State, police interrogated a thirteen year old who was a murder suspect.72 Here is how the court described the waiver:
Detective Knowles stated that, because of appellant’s age, he employed a different procedure when he advised her of her rights. Although, when questioning an adult, Detective Knowles would normally explain the Miranda rights one time only, because of appellant’s age he went through the Miranda rights twice. He stated that appellant verbally and physically indicated that she understood the rights, and that he was convinced she understood the waiver when she signed it. … Detective Knowles further stated that appellant appeared to be alert and oriented, that her demeanor was very cool and calm, and that her speech was clear and forthright.73
The Arkansas Court of Appeals rejected the waiver argument in a single paragraph, which concludes: “Although appellant’s age is a factor to consider in determining the voluntariness of her waiver, our review of the totality of the circumstances has not convinced us that the trial court’s ruling on the issue of waiver was clearly erroneous.”74 At least the Arkansas court, unlike the South Carolina court, recognized that waiver is the critical issue when a very young suspect is interrogated. But we are not persuaded that a valid waiver is proven by reading the warnings twice to a “cool and calm” thirteen year old. In our view, it adds nothing for the officer to make the self-serving statement that he believed that she understood her rights.
The Matthews court relied on an Arkansas case holding that a twelve year old who read at a fourth-grade level waived his Miranda rights.75 An “educational examiner,” apparently the only expert witness in the earlier case, testified that the boy lacked the “intellectual development … to perceive the implications and consequences of signing [the] form after having it read to him.” So what did the Arkansas court rely on? The juvenile and his mother, who was present, “agreed that his statement was not coerced,” and the juvenile was “calm and attentive” during the three and one-half hour interrogation with no break.76 But a suspect’s demeanor and the absence of coercion are both irrelevant to a juvenile’s capacity to understand and waive Miranda. The issue should be capacity, not whether his will was overborne or how he appeared.
In re M.D. involved a thirteen year old suspected of vandalism.77 His “waiver” was secured after he was taken to the sheriff’s office where
Deputy Perry read him his Miranda rights. Deputy Perry explained to appellant that he had the right to an attorney, and had him read over a Miranda waiver. Deputy Perry testified that appellant (1) then signed the waiver “free [sic] and willingly,” (2) showed remorse after writing his confession, and (3) never asked questions about the waiver. Appellant never asked for an attorney or his parents.78
One might have expected the court at least to concede that this is thin waiver gruel when the suspect is thirteen. The officer read the rights to him, told him he had a right to a lawyer, and then had him read the waiver form. That the officer believed the waiver valid, that the juvenile showed remorse after confessing, and that he never asked questions about the waiver or requested help are completely beside the point in analyzing waiver. But the Ohio Court of Appeals was satisfied. The court cited these facts and added two other irrelevant facts: that one of the officers was “comfortable appellant understood what a Miranda warning was” and that “there is nothing in the record to suggest he was of insufficient intelligence based upon his age or that he was mentally impaired and was unable to understand his Miranda rights.”79 The court thus held that a thirteen year old, of ordinary intelligence for that age, can understand, and act on, his Miranda rights without any attempt to explain the complex, underlying concepts.
Even when a juvenile suspect wins, it feels like a loss. Wesley was “an educationally impaired youth of low average intelligence” and eleven years old when interrogated.80 The officer noted that “Wesley was ‘slow,’ fidgety, and had trouble paying attention.” When Wesley’s attention wandered, “the officer would try to ‘get him back to my attention.’” The court-appointed psychologist found “that Wesley was ‘very immature’ and had reading abilities of the second grade, fourth month.” He also testified during the suppression hearing that “in his opinion, it was beyond the cognitive abilities of most eleven-year-old children to make a free and unconstrained choice about whether to speak to a police officer”—a position we very much embrace—“and that Wesley’s impairments rendered it substantially more difficult for him to decide whether or not to confess.”
The officer interrogated Wesley for almost two hours. He testified it was a “constant struggle to keep Wesley focused on him and his questions.”81 Yet the trial judge admitted the confession, and the New Hampshire Supreme Court retreated to a laundry list of factors to hold that the confession was involuntary: “The repeated and prolonged questioning of eleven-year-old Wesley, despite his continued denial of any involvement in the fire, coupled with the absence of any supportive adult during the interview, lead us to conclude that Wesley’s will was overborne.”82 But the court appears to rely largely on the New Hampshire rule that the State had to prove voluntariness “beyond a reasonable doubt,” noting that the evidence presented to the trial court is perhaps “sufficient to support a finding by a preponderance of the evidence” in favor of admitting the confession.83 The federal standard is preponderance of the evidence.84 If most states followed that standard, then Wesley’s confession might have been admissible despite his age and reading level!
We would have written a different opinion, holding that an eleven year old who has the reading ability of a seven year old cannot waive Miranda, by any standard of proof, and thus cannot be interrogated. That is the message we want to send the police. To be sure, Wesley apparently committed the serious crime of trying to burn down his house. But if the authorities cannot prove it without his confession, we would be happy to allow civil commitment to proceed rather than a delinquency adjudication.
The net effect of these cases is to put the burden of proof on very young suspects to demonstrate they are of “insufficient intelligence” or “mentally impaired,” rather than putting the burden of proving waiver on the State. Most courts effectively put the burden of disproving waiver on adult defendants. While this is a retreat from language in Miranda, it can be justified because the Court has morphed Miranda into a notice case. But how can courts believe that eleven and twelve year olds are effectively put on notice by rights that many adults cannot comprehend?
Thomas Grisso’s study found that 63% of juveniles and 37% of adults received a “zero” credit on one or more items in terms of comprehending the terms in the warnings.85 When the issue was understanding the significance of the rights, 45% of juveniles, compared with 15% of adults, received zero credit for understanding the right to counsel. “The most common confusion about this warning statement concerned the time and place an attorney could be consulted, ‘interrogation’ often being misconstrued as an adjudication hearing.”
Though most juveniles understood that they did not have to talk to police, 55% “believed that they would have to explain their criminal involvement in court if questioned by a judge.”86 The ability to comprehend and understand the warnings varied, as one would expect, by age. Grisso concluded that “juveniles younger than fifteen years of age failed to meet both the absolute and relative (adult norm) standards for comprehension.”87 The “vast majority of these juveniles misunderstood at least one of the four standard Miranda statements, and compared with adults, demonstrated significantly poorer comprehension of the nature and significance of the Miranda rights.”
Barry Feld examined tapes of interrogations of juveniles by the St. Paul police department, limited to those who were sixteen and seventeen years old.88 His conclusions are consistent with Grisso’s field research. Feld found, “Sixteen- and seventeen-year old juveniles exhibit relatively adult-like competence in the interrogation room … [and] do appear to understand the contents of the Miranda warnings.”89 Indeed, “about the same proportions of [sixteen- and seventeen-year old] juveniles as adults appear able to exercise their rights both prior to and during the course of questioning.” The significance of Grisso’s and Feld’s findings is to support the Haley Court’s intuition that “a boy of fifteen, without aid of counsel,” would not “have a full appreciation of” the right to remain silent.90 The Court was also right to reject the confession of the fourteen-year-old suspect in Gallegos and, we believe, the state courts that found valid waivers from children aged eleven, twelve, and thirteen are wrong as a matter of law.
Grisso and Feld support the Fare v. Michael C. Court’s assumption that a juvenile who is sixteen years old is capable of waiving Miranda. These studies do not, however, support the Court’s easy assumption that Michael C. did waive his rights. There is, of course, no magic age when a juvenile suddenly becomes capable of waiving Miranda rights. What is needed is a more searching inquiry into the juvenile’s capacity, much like the Court did in the due process cases that preceded Miranda. Certainly, more is required than merely finding that a juvenile was of average intelligence. The findings of Grisso and Feld manifestly reject as absurd cases applying the Michael C. analytical framework to suspects who are eleven, twelve, and thirteen.
We do not mean to suggest that all courts treat young suspects as having adult capacity to waive Miranda and make voluntary confessions. A counterexample is a case arising out of the murders of nine people, including six Buddhist monks, in a temple near Phoenix in 1991.91 Investigators initially arrested and interrogated four individuals from Tucson. After relentless questioning, all made incriminating statements. They later recanted their confessions, and the authorities ultimately agreed that the confessions were false. A lawsuit against the Maricopa County Sheriff’s office was settled in 1994 for $2.8 million.
Authorities later focused on Johnathan Doody, seventeen years old, who was interrogated from nine at night until ten the next morning.92 The interrogator’s “recitation of Miranda’s basic warnings consume twelve pages of transcript, largely a byproduct of the detective’s continuous usage of qualifying language.” In an en banc ruling, the Ninth Circuit concluded that the Miranda warnings, as given, “negated the intended effect of the Miranda warning.”93 The court also held that Doody’s confession was involuntary.94 The Supreme Court vacated the Ninth Circuit’s judgment and remanded for reconsideration of the Miranda holding in light of Florida v. Powell95 but, presumably, the judgment about voluntariness will stand. Whether it stands, the opinion strikes the right tone and asks the right questions about how juveniles might perceive a relentless police interrogation.
The Ninth Circuit began its due process voluntariness analysis by quoting Justice Frankfurter’s opinion in the 1949 case of Watts v. Indiana:
When a suspect speaks because he is overborne, it is immaterial whether he has been subjected to a physical or a mental ordeal. Eventual yielding to questioning under such circumstances is plainly the product of the suction process of interrogation and therefore the reverse of voluntary. We would have to shut our minds to the plain significance of what here transpired to deny that this was a calculated endeavor to secure a confession through the pressure of unrelenting interrogation. The very relentlessness of such interrogation implies that it is better for the prisoner to answer than to persist in the refusal of disclosure which is his constitutional right.96
The Ninth Circuit noted that the United States Supreme Court “has consistently reminded us that ‘admissions and confessions of juveniles require special caution.’”97 But the most recent Supreme Court citation that the majority offers is the 1967 case of In re Gault. Indeed, most of the Supreme Court quotes came from the 1940s, the 1950s, and the early 1960s—the heyday of the due process test. Naturally, the Ninth Circuit focused on Gallegos and Haley, cases that essentially held that suspects who were fourteen and fifteen simply were not capable of understanding the consequences of cooperating with police during intense interrogation.
To be sure, Doody was seventeen, not fourteen. Thus, critical to the majority’s analysis were the aspects of the nearly thirteen-hour interrogation that were designed to sap Doody’s will to resist: (1) He was an “unsophisticated teenager”;98 (2) he was not offered a break during the first nine hours of the interrogation;99 (3) he was not a “native of this country”; (4) he was interrogated by “tag-teams of two, three and four detectives” whose tone ranged from “pleading” to “scolding to sarcastic to demeaning to demanding” to “downright chilling”;100 (5) almost thirteen hours of interrogation took place overnight, making him “sleep-deprived”;101 and (6) immediately before he confessed, the detectives told him “that he had to answer their questions.”102 While it might not be as relevant as the factors just mentioned, the court also noted, “By the end of the interrogation, Doody was sobbing almost hysterically.”103 Moreover, “this same task force questioned four adult men and, undoubtedly using the same tactics, procured what the State concedes were false confessions from all four.”104 If “the will of four adult men was overborne to the extent that they confessed to murders they did not commit,” why wouldn’t “the will of this young teen” be “similarly overborne”?
Luckily for Doody, the police taped the interrogation—seventeen audiotapes in all.105 It seems unlikely that Doody would have won, even in the liberal Ninth Circuit, if all the defense had was his recollection and the testimony of the interrogators. “The audiotapes of Doody’s interrogation are dispositive in this case, as we are not consigned to an evaluation of a cold record, or limited to reliance on the detectives’ testimony.”106 Instead, the court heard “an extraordinarily lengthy interrogation of a sleep-deprived and unresponsive juvenile under relentless questioning for nearly thirteen hours by a tag team of detectives, without the presence of an attorney, and without the protections of proper Miranda warnings.” The tapes “reveal a picture that bears no resemblance to the avuncular scene painted by the Arizona Court of Appeals.”107 Nor was the Doody majority impressed with the evidence of maturity on which the state courts relied—his “participation in ROTC, his not-yet-completed high school studies, his work as a grocery store bagger, his ability to speak English as a second language and his lack of mental disability.”108
The Ninth Circuit adopted an approach to analyzing the coercive factors that is very different from that used by the Arizona Court of Appeals and by most courts today. Doody acknowledged the cumulative effect of the various factors rather than analyzing each factor in isolation. As we will see in more detail in the next section, the Ninth Circuit’s approach is the one the Supreme Court used before Miranda.
Three of the eleven judges who comprised the en bane panel in Doody found the thirteen-hour interrogation to comport with due process—or at least close enough that the deference owed by federal courts when reviewing final state court decisions should have carried the day.109 Whatever the right result when applying a deferential standard of review, we believe that the trial court and state court of appeals reached the wrong result when they found that neither Miranda nor the due process clause was violated by the interrogation that Doody faced.
And we turn, now, to due process claims more generally.
As Leo put it in 2008, “The Miranda warning and waiver ritual. … . makes almost no difference in American police interrogation because virtually all suspects waive (or are legally constructed to have waived) their Miranda rights, and almost no confession is ever excluded from evidence at trial because of a Miranda violation.”110 For the large universe of suspects who waive their rights, the Miranda “revolution” did not protect them from police interrogation. Instead, Miranda has effectively deprived those suspects of the voluntariness flourish that the FDR Court created from the due process clause. Thus, odd though it may sound, and as totally weird as it would have sounded to the nine members of the Miranda Court, a case can be made that suspects as a whole are worse off now when they get to court than before Miranda was decided.
A harder question is whether suspects are also worse off in the police interrogation room. At one level, the answer is likely “no.” As Leo has observed, Miranda has almost certainly contributed to a “civilizing” of police interrogation practices.111 The third degree was on its way out when Miranda was decided but, according to interrogation manuals, police interrogators still seemed to treat suspects as prey. By the 1980s, the “training manuals taught police a kinder and gentler way to interrogate while still creating the opportunity to overbear the will of suspects.” Precisely how much Miranda contributed to this “civilizing” of police is impossible to know. But being trained to tell your adversary that he need not cooperate with you and that you must accept his decision not to talk or to consult with counsel must, over time, contribute to that “kinder and gentler” police interrogation. Whether that is a benefit to suspects is a difficult call. Confessing after having one’s will worn down in a kindly way seems just as detrimental to suspects, if more pleasant, as earlier interrogation methods.
In chapter 7 we surveyed some of the key voluntariness cases as the Court made its way, in fits and starts, toward Miranda. The focus showed how the cases built toward Miranda. In this chapter, we want to show how the due process doctrine can be understood on its own terms and without regard to the path to Miranda. We begin that task in this subsection by showing one of the chief virtues of the voluntariness test—its flexibility. We offer two examples.
In Fikes v. Alabama, a black man was arrested as a suspect in a series of burglaries and rapes.112 In examining whether the interrogation violated the due process clause, the Court stressed: “It is, of course, highly material to the question before this Court to ascertain petitioner’s character and background.”113 The Court continued:
He is a Negro, 27 years old in 1953, who started school at age eight and left at 16 while still in the third grade. There was testimony by three psychiatrists at the trial, in connection with a pleaded defense of insanity, to the effect that petitioner is a schizophrenic and highly suggestible. His mother testified that he had always been “thick-headed.” Petitioner worked in a gas station in his home town of Marion, some 30 miles from Selma. So far as appears, his only prior involvement with the law was a conviction for burglary of a store in November 1949; he was released on parole in January 1951.
Fikes was interrogated repeatedly for about ten days, but there was no evidence of coercion or physical abuse.114 Yet focusing on his “low mentality,” his “isolation,” and the failure to take him before a magistrate as required under state law, the Court held that the “totality of the circumstances that preceded the confessions in this case goes beyond the allowable limits. The use of the confessions secured in this setting was a denial of due process.”115 The flexibility and indeterminacy of the pre-Miranda due process test is on full display in Fikes.
A virtue of Miranda from an efficiency perspective is that it treats all suspects alike. Anyone, even someone who is mentally disabled or eleven years old, is presumed able to understand that “you have the right to remain silent” means one does not have to answer police questions. It is, we have tried to point out, a highly contestable assumption and turns out to be a vice from the perspective of justice or fairness. The due process “voluntariness” test, on the other hand, was horribly inefficient because it is based on a “complex of values” and “the role played by each in any situation varies according to the particular circumstances of the case.”116 In Haley v. Ohio, where the suspect was fifteen, the Court noted, “Mature men possibly might stand the ordeal from midnight to 5 a.m. But we cannot believe that a lad of tender years is a match for the police in such a contest. He needs counsel and support if he is not to become the victim first of fear, then of panic.”117 Voluntariness is flexible enough to permit a court to find the confessions of Fikes and Haley a violation of due process even if the interrogation would not have violated the due process rights of white businessmen like Lisenba or Ashcraft.
So when 1966 dawned, the Court’s voluntariness doctrine was supple and fact-sensitive. But the very flexibility that was the virtue of the Court’s voluntariness test in the hands of the Supreme Court permitted lower courts to admit confessions pretty much whenever the judge thought it made sense. Trial and state appellate courts, after all, had found no reason to suppress the confessions of Haley, Gallegos, and Fikes, along with eight other state cases that the Court reversed in the period 1957 to 1963—an average of more than one a year.118 One reason to decide Miranda as if it were a statute regulating interrogation was to get the Court (and the lower courts) out of the business of reviewing thousands of voluntariness claims. It worked, though probably not in the way the Court imagined. The Miranda opinion contemplates a world where lawyers guide suspects through the minefield of police interrogation, where uncooperative suspects insist on their right to remain silent, or where contrite suspects volunteer a confession.
Instead, most suspects waive Miranda and courts accept acts of waiver as strongly presumptive of voluntariness, both of the waiver itself and of statements made following waiver. In the Court’s words, only “rare” cases permit “a colorable argument that a self-incriminating statement was ‘compelled’ despite the fact that the law enforcement authorities adhered to the dictates of Miranda”.119 An empirical study of all lower court Miranda cases in 1999 and 2000 bears out how “rare” these cases are. Welsh White found only nine cases that held the postwaiver confession involuntary, and four of those were based on state constitutional law.120
It is possible that Miranda has made the due process voluntariness test even less protective than it was prior to 1966. Courts sometimes use the warnings “offensively” to sanitize the interrogation that follows by relying on the fact that warnings were given repeatedly. Proving waiver sometimes has the effect of eliding or suppressing the question of whether the subsequent statements are voluntary. In the garden-variety case, these developments make sense as a starting place. If there is doubt about whether the suspect answered police questions voluntarily, the repeated warning that he need not answer should weigh in favor of voluntariness. And if the statement follows relatively quickly after waiver, this, too, is evidence of voluntariness. In many cases, however, courts allow the warnings to carry more weight in the voluntariness calculus than we think justified.
In his analysis of Chavez v. State, Alfredo Garcia provides a dramatic example of the “offensive” use of Miranda.121 Detectives in three shifts interrogated Chavez for thirty hours with minimal breaks.122 After thirty hours, he was allowed to sleep for six hours—on the floor of the police interview room. He was then interrogated for another twelve hours. Twice during the marathon interrogation, a polygraph examiner told him that he had failed the polygraph test.123 The Florida Supreme Court in a per curiam opinion, with not a single dissenting justice, held that Chavez’s interrogation produced a voluntary confession.124
What is more disappointing than the outcome is the court’s analysis. As for a Miranda waiver, Chavez’s initials on a Spanish-language rights form apparently provided all the evidence the court needed to conclude that he had “knowingly” waived Miranda.125 As for voluntariness of his statements, what follows is the state supreme court’s analysis of whether Chavez’s confession was voluntary:
Although Chavez was questioned over the course of several days, he was provided with food, drink, and cigarettes (as requested) at appropriate times, and permitted to have frequent breaks. His interrogation was also interspersed with time away from the police facilities for visits to various properties, a six-hour rest period (where Chavez was offered a blanket and a pillow), and times when he was left alone for quiet reflection. He was repeatedly given Miranda warnings, in Spanish, and indicated each time that he fully understood them. Consequently, the trial court did not err in denying Chavez’s motion to suppress on this ground.126
Notice how courts manipulate facts to get the result they want. The hours spent on the floor between interrogations were a rest period. The court never mentions that he had to sleep on the floor—this comes from Garcia’s study of the trial record127—but the court seems to think it important that he was offered a blanket and a pillow, as if this somehow reduced the pressure of the interrogation. The police twice took him to the scene of the crime while questioning him about where he hid the body. The court viewed these trips a respite: “time away from the police facilities for visits to various properties,” as if being brought to the scene of the crime while being interrogated was a pleasure trip. What was the most important indicator of voluntariness? “He was repeatedly given Miranda warnings, in Spanish, and indicated each time that he fully understood them.”
As Garcia concluded, “Rather than assisting the suspect, what Miranda wrought for Chavez was an argument that a seventy-five hour interrogation which resulted in various incriminating statements was voluntary and thus constitutional.”128 Garcia then asks us to
[i]magine for a moment a world without Miranda. Without the waivers, would the prosecution and the trial court have found it more difficult to establish that Chavez’s incriminating statements were “voluntarily” given? Does not the case seem redolent of Ashcraft v. Tennessee, where the Court held that a thirty-six hour continuous interrogation of the suspect was “inherently coercive” and not voluntary?. . . Quite simply, what the officers had in Chavez that the officers in Ashcraft lacked was the great Miranda warnings to sanitize a clearly involuntary confession.
Garcia is correct that the Ashcraft Court would have decided in Chavez’s favor. But what he misses is that the Tennessee Supreme Court, with one dissenting justice, had ruled against Ashcraft.129 Reading Lisenba to “concede[] to each state the right to determine the question of the admissibility of confessions to itself, subject only to the overpowering requirement … of fair play,” the state court thought Ashcraft insufficiently noteworthy even to publish the opinion.130 It was only by virtue of the United States Supreme Court granting certiorari that Ashcraft won his involuntariness claim.
The FDR-flavored Court that decided Ashcraft’s case was probably uniquely positioned to shift the emphasis from whether the suspect’s will was overborne—Lisenba—to whether unacceptable pressures were placed on him—Chambers and Ashcraft. It is not as clear to us as it is to Garcia and Welsh White131 that the due process voluntariness analysis applied today by lower courts is meaningfully different than it was pre-Miranda. To be sure, courts today sometimes rhetorically rely on the warnings to “sanitize” a high-pressure interrogation. The Eleventh Circuit, for example, rejected the defendant’s involuntariness argument in Morgan v. Zant as “without merit,” without even disclosing the length of the interrogation, because the officers testified “that Morgan received numerous Miranda warnings prior to making any statement, and that he willingly confessed, these warnings notwithstanding.”132
Giving Miranda warnings is, of course, relevant to voluntariness. Chavez and Morgan could have ended the interrogation at any point by simply invoking their rights. Repeated warnings reminded them that they did not have to talk. And lower courts did not need Miranda warnings to find admissible confessions made in cases against Fikes, Blackburn, Ashcraft, White, Haley, Gallegos, Spano, Chambers and other defendants. Thus, it appears that not much has changed in due process voluntariness in the lower courts.
Judges elide the issue of voluntariness of statements with the waiver question for two sensible reasons. Judging from judicial opinions, defense lawyers do not distinguish the question very often. And the two questions are governed by the same standard of voluntariness. Moreover, the Court in 1984 said that Miranda’s safeguards are intended “as much as possible to free courts from the task of scrutinizing individual cases to try to determine, after the fact, whether particular confessions were voluntary.”133 As the New Hampshire Supreme Court recently put it: “While compliance with Miranda does not conclusively establish the voluntariness of a later confession, it is one factor that a trial court can consider.”134
But courts sometimes confuse waiver and voluntariness. In State v. Reliford, the Tennessee Court of Criminal Appeals introduced the suppression issue by saying that the “trial court determined that the defendant signed the advice of rights form and that he freely and voluntarily talked to the police officers.”135 This sounds like the court saw the issue as the voluntariness of the confession. The next sentence is, “The trial court’s determination at the suppression hearing that a confession was voluntary is presumptively correct on appeal.” The issue now clearly seems the voluntariness of the confession. Two sentences later, the court wrote, “When determining whether an accused has voluntarily, knowingly, and intelligently waived his Miranda rights, this court must consider the totality of the circumstances which existed when the accused waived these rights.” Now the focus is on waiver.
When resolving the issue in favor of the State, the court kept the emphasis on waiver: “The trial court accredited the testimony of the sergeant and determined that the defendant knowingly and voluntarily signed the waiver of rights form.” Of course, it is possible that the trial court expressly resolved both the waiver and voluntariness issue. If so, the appellate court opinion obscures that clarity.
As long as courts can move back and forth effortlessly between the waiver question and the voluntariness question, they can obscure and avoid most due process issues.
One way to find a confession admissible is to put a heavy burden on defendants to demonstrate significant coercion before finding a due process violation. To make the defendant’s task even more difficult, courts often examine each objectionable technique in isolation and then ask whether it is, by itself, coercive. Lower courts rarely do what the Court did in the 1940s to the 1960s, which is to look at the “totality of the circumstances.” To be sure, courts sometimes say that is what they are doing. But the proof is in the method by which the totality is analyzed.
For example, in State v. Harris, the suspect was handcuffed and taken to the police station where he was placed in a 10 foot by 10 foot interview room and “shackled to the floor.”136 Except for bathroom breaks, he remained shackled to the floor for seven hours and was interrogated for two and one-half hours. The officers suggested to him that witnesses had seen him at the scene of the shooting, that witnesses had picked his photo out of a photo array, and that there was blood on his coat.137 All of these incriminating facts were false. Although eyewitnesses had picked photos of other men from the array, the officers did not mention that inconvenient fact. Attempting to provide exculpatory evidence, Harris asked to use the phone to call an alibi witness.138 The police refused the request. As for Miranda, the court stated that the warnings were given and he “agreed to talk” to the police.139
Harris, unlike Chavez, does not rely on the Miranda waiver in resolving the due process inquiry. Instead, the court looked at each potentially coercive fact in isolation and placed the burden on Harris to show that it rendered his confession involuntary: “Harris fails to cite any case law establishing that a nearly 7-hour detention is too long. Likewise he has not pointed to any case law to support his proposition that his confession was involuntary because he was restrained by shackles.”140 Recall that eight hours of questioning plus the “false friend” technique caused Spano’s will to be “overborne by official pressure, fatigue and sympathy falsely aroused.”141 If the same Court had been writing the Harris opinion, it would have gone something like this: “Shackled to the floor for seven hours, led to believe by police lies that his case was hopeless, facing interrogators whose purpose was to break his will, denied the opportunity to speak to the outside world, Harris saw no choice but to confess. Moreover, police who intentionally mislead a suspect about evidence of guilt are acting in a way that is inconsistent with due process. In sum, he was ‘subjected to pressures to which, under our accusatorial system, an accused should not be subjected.’142 The confession must be suppressed.” But the Kansas Supreme Court unanimously held Harris’s confession voluntary and thus admissible.
In State v. Jackson, Raleigh police told Jackson “that they wanted to talk with him as a possible witness” in a murder case.143 Though not under arrest, he accompanied police to the Investigative Division where he “was fully advised of his Miranda rights and waived those rights in writing.” The North Carolina Supreme Court remarked, “He never requested an attorney.” He was “interviewed” for almost three hours and agreed to return the next day “for further interviews and to take a polygraph test.” The detectives also asked for the shoes and pants he had worn the day of the homicide. He gave them shoes and pants but not the ones he had worn that day.
The next day, he again waived Miranda in writing and took a polygraph.144 The state court tells us, again, that he “did not request an attorney at any time.” He was “interviewed” for four hours after the polygraph. A detective told him that there was blood on his shoes and pants, though Jackson, of course, knew that could not be the victim’s blood because he gave the police the wrong clothes. When the “interview” was over, police told Jackson “that they would get back in touch with him later.”
About twelve days later, police approached Jackson on the street and told him that they “wanted to talk with him about the murder.”145 He was not arrested and, the state court assured the reader, “voluntarily got into the car with Officer Williams and went to the police station.” By now, the routine should be familiar: Jackson “was again advised of all of his Miranda rights, waived those rights in writing, and consented to be questioned by the officers without the presence of an attorney.” Police “told the defendant he was not under arrest and that he was not in custody. He was not handcuffed and the interview room door was not locked.”
The “interview” lasted four and one-half hours, during which police lied again about the evidence.146 The detective showed him a knife with a bloody fingerprint that the police said was Jackson’s. He told Jackson that his fingerprints were also found in the victim’s apartment and that an eyewitness could identify him leaving the victim’s apartment with that knife. The state court conceded that “[t]hese statements were not true.” The detective told Jackson that the maximum penalty for first degree murder was the death penalty. But the state court assured the reader that “defendant was not threatened that if he did not cooperate he would get the death penalty.” At some point, one of the officers said that if he pleaded not guilty “that the other officers would probably go into court and testify that defendant was a black man, killing and raping white women.”147
Just before midnight, one of the detectives
“. . .went over the evidence again with the defendant, stating that the police had a murder weapon, they had defendant’s fingerprints and the defendant’s fingerprints at the scene, that they were on a knife sharpener and on a wooden post on the decedent’s front porch, and that they had a witness who saw the defendant coming out the door carrying a knife. These statements were not true. At that point the defendant responded that he did it.148
The state supreme court held, 4–3, that Jackson’s confession was voluntary. The key facts, in the court’s view, that support a finding of voluntariness were: (1) He “was never in custody or under arrest before he confessed”; (2) an officer told him “on the day he confessed that he was free to go at any time”; (3) “he was not restrained, not touched, threatened, or intimidated”; (4) “he was taken where he wanted to go after the first two sessions with the officers”; (5) “he walked to the police station one day”; (6) “there was at least a week between the second and third interviews”; (7) Jackson “had an extensive criminal history and had previous experience with interrogation”; (8) “Jackson was repeatedly given proper Miranda instructions although he was not in custody”; (9) Jackson “was not questioned for undue periods of time, and no promises or threats were made to him”; (10) Jackson “made misrepresentations to the officers and was aware to some extent that the officers were not truthful with him.”149
This list is less impressive than its length might suggest. Facts (l), (2), (3), (4), (5), and (8) simply restate the same theme: Jackson was not physically restrained and had been repeatedly given warnings. Presumably, the court is inviting the reader to infer that Jackson was not in custody and thus no Miranda violation was even possible. In addition, if Jackson believed he was free to leave, it would weaken his coercion argument. The time between interviews, (6), seems irrelevant to us. That he had prior experience with interrogation is relevant but not, in our view, particularly weighty. As for his knowledge that the police had lied to him about blood on his pants and shoes, we believe that should be added to the coercion side of the balance, as we will explain shortly. Thus, almost the entire “work” in the court’s voluntariness calculation is done by the lack of custody and fact (9) that he “was not questioned for undue periods of time, and no promises or threats were made to him.”
But the lack of promises or threats is contestable. The dissent added some facts that the majority conveniently left out: The police told Jackson that “no one could help him after he was placed in the gas chamber”; that “if his girlfriend was bearing his child he would not be the one to raise it if he were convicted”; “that if there were extenuating circumstances, defendant should ‘bring [them] to light’ or otherwise risk suffering the death penalty”; that Jackson “might avoid the death penalty by playing on the jurors’ emotions at trial”; that “if he told the truth ‘it would certainly come out in court that he cooperated.’”150
Then there was the “promise” that if he pleaded not guilty the other officers would go to court and testify that he was a “black man, killing and raping white women.” No threats? No promises? One should keep in mind the context—Jackson was a black man in North Carolina; the victim was a white woman; and North Carolina has a history of racial violence. North Carolina was, of course, a slave state that joined the Confederacy. It saw the first protest in the nation against white-only lunch counters, at the F.W. Woolworth lunch counter in Greensboro in 1960. The protests prompted deadly white backlash.151 About eighteen months before police “interviewed” Jackson, marchers in a “Death to the Klan” rally in Greensboro were gunned down by group of Ku Klux Klansmen and members of the American Nazi Party.152 Five marchers were killed. Greensboro is eighty miles from Raleigh.
An all-white state jury acquitted the Klansmen and Nazis six months before Jackson confessed to the murder in Raleigh. When Jackson was repeatedly questioned by the Raleigh police, his world included the history of slavery and segregation, the deaths in Greensboro, and the acquittals of the Klansmen and Nazis who killed the marchers. The police aggravated Jackson’s vulnerability by telling him that he faced the gas chamber, they knew he was lying to them, and they would testify in court about his vicious murder and rape of white women. Jackson would have known the effect on the jury of the murder and rape accusation. For shameful decade after shameful decade, black men charged with raping white women had been lynched—guilty or not—a fact that Jackson surely knew.
Moreover, he faced what appeared to be overwhelming evidence against him, constructed out of police lies. The majority seemed to believe that his knowledge of the police lying about some evidence might lead him to think that they were lying about other evidence. But the opposite inference is, to us, more likely. A black man charged with raping and murdering a white woman in North Carolina in 1981 could very easily believe that the police would fabricate whatever evidence it took to send him to the gas chamber. We prefer the dissent’s reading of Jackson’s interrogation: “Absent torture or other physical abuse, it would be difficult to conceive of interrogation tactics more likely to produce an untruthful, unreliable confession than the ones utilized in this case.”153
Many scholars believe that the problems we have surveyed here are the result of later Courts and lower courts failing to apply Miranda’s principles as robustly as the Warren Court intended. But we must not forget Seidman’s powerful critique from the last chapter. Perhaps Miranda was, from the beginning, a compromise, and a retreat from a world in which lawyers attended all interrogations. Perhaps Miranda all along was meant to support the status quo where police interrogated mostly guilty suspects and usually obtained a confession. Indeed, according to Seidman, Miranda works better to maintain the status quo than the voluntariness test. “The imagery of consent and equality, so powerfully and simply invoked by Chief Justice Warren’s rhetoric, dominates popular culture in a way that the old confusing, complex, and contradictory law of … voluntariness never could.”154 In sum, Miranda has “led us into a trap” and the “escape route from our present predicament is not nearly so well marked” as it was before Miranda.155
No one knows how many suspects are moved to stay silent because of the warnings. Nor do we know how many suspects might talk because of the warnings. They are, after all, an official indication that the police believe the suspect is guilty, and they thus invite a response if the suspect is to maintain that he is innocent. We do know that roughly 80% of suspects waive their rights and speak to the police. After waiver, they face the inherent compulsion that Miranda condemned. We agree with Bradley that Miranda “has surely not wiped out confessions that due to factors like physical coercion, psychological ploys, mental incapacity, and fatigue are not sufficiently ‘voluntary’ to be considered either fair or reliable.”156
We would partly agree with Garcia that “Miranda has produced paradoxes that the Warren Court would have never imagined—circumstances in which the ‘old’ voluntariness approach affords more protection to criminal suspects than the putative safeguards outlined in the decision.”157 Garcia’s paradoxes are there as long as the old due process approach is that of the Supreme Court of the 1940s, 1950s, and 1960s. Given the repeated need to reverse state courts that found confessions voluntary on gruesome facts, it strikes us as unlikely that lower courts were embracing the Supreme Court’s robust due process approach. But, at a minimum, defendants today are deprived of the lofty rhetoric of the FDR Supreme Court. The Ashcraft-Spano due process protection seems as far as a distant star.
Sandy Thompson despairs that “there is truly nothing left of Miranda. … The Miranda chapter has effectively come to a close.”158 Garcia would have the Court overrule Miranda. But we do not despair as much as Thompson or Garcia for three reasons. First, Dickerson v. United States affirmed Miranda in 2000,159 thus effectively closing the door on overruling it, at least for many years (Garcia wrote two years before Dickerson was decided). Second, viewing Miranda as due process notice makes clear its appeal: It is part of basic fairness to tell a suspect that he has a right not to answer police questions, and that part of the Miranda chapter continues to resonate. Third, we hope our solutions presented in chapter 10 will move courts, at least partly, back to the Supreme Court’s robust due process voluntariness doctrine that existed before it was eclipsed by the glare of Miranda.