Sample Characteristics

As indicated in Table 1, males comprised the vast majority (89.3 percent) of the 307 youths whom police questioned. Prosecutors charged more than half (55.0 percent) with property offenses—burglary, larceny, and auto-theft. They charged nearly one-third (31.6 percent) of youths with crimes against person—murder, armed robbery, aggravated assault, and criminal sexual conduct. They charged the remaining youths with drug crimes (6.2 percent), firearm offenses (5.5 percent), and other felonies (1.6 percent). Prosecutors charged more than half (56.4 percent) with only one felony, an additional quarter (25.1 percent) with two crimes, and the remainder with three or more. The group lacks some of the most serious offenders because prosecutors filed certification motions and juvenile court judges transferred them to criminal court.

Table 1. Characteristics of Juveniles Interrogated

^a Crimes against property include: burglary, theft of a motor vehicle, arson, receiving stolen property, possession of stolen property, possession of burglary tools, criminal damage to property, theft, forgery, theft by swindle, and credit card fraud.

^b Crimes against the person include: aggravated and simple robbery, aggravated assault, murder and attempted murder, criminal vehicular homicide, criminal sexual conduct, and terroristic threats.

^c Drug crimes include: sale or possession of a controlled substance—crack, methamphetamine, marijuana, codeine, ecstasy, heroin—possession of a forged prescription, and tampering with anhydrous ammonia equipment (methamphetamine).

^d Firearm crimes include: possession of a firearm, discharge of a firearm, theft of a firearm, possession of an explosive device, and drive-by shooting.

^e Other offenses are fleeing a police officer.

Nearly one-third (30.6 percent) of juveniles had no prior arrests. Police previously had taken into custody more than one-third of these youths for non-criminal status offenses (15.3 percent) or misdemeanors (22.8 percent). About one-third of these youths (35.1 percent) had one or more prior felony arrests and more than half (57 percent) had prior juvenile court referrals. Nearly one-third (29.9 percent) were under court supervision—probation, placement, or parole status—when police questioned them. About half of the youths were white (52.1 percent) and the remainder (47.9 percent) members of ethnic and racial minority groups—Black, Hispanic, Native American, and Asian. Black juveniles accounted for more than one-third (34.9 percent) of the sample. Compared with these counties' 16- and 17-year-old felony caseloads, the interrogation group included a larger proportion of males, more youths charged with property and violent crimes and with prior court referrals, and fewer charged with drug offenses (Reference Feld and SchaeferFeld & Schaefer 2010a, Reference Feld and Schaefer2010b).

Securing Miranda Waivers

When police take suspects into custody and interrogate them, Miranda requires officers to warn them to dispel the inherent coercion of isolation and questioning. Police had formally arrested the vast majority (86.6 percent) of these juveniles prior to questioning. They made a Scales recording of each interrogation whether they initially arrested or later released a youth. Police detained nearly two-thirds (61.7 percent) of those whom they questioned and released the others to parents. More than half (55.7 percent) of interrogations took place in police stations. Another quarter (23.1 percent) occurred at juvenile detention centers. Thus, police questioned more than three-quarters (78.8 percent) of youths in interrogation rooms. Nearly one-tenth (8.1 percent) of interrogations took place in a police car at the place of arrest. Police conducted 6.2 percent of interrogations at juveniles' homes and another 6.2 percent in schools. Every juvenile in the sample received a proper Miranda warning and one-fifth (19.5 percent) of the files contained an initialed and signed warning form.

Although Miranda requires police to warn suspects, officers' goal to solve crimes provides no incentive to encourage them to invoke their rights. This inherent contradiction requires officers to engage in a quasi-confidence game—“systematic use of deception, manipulation, and the betrayal of trust in the process of eliciting a suspect's confession” (Reference LeoLeo 1996b:259). Police used several tactics to predispose suspects to waive Miranda without alerting them to its significance—admonishing them to tell the truth, minimizing the warning, or advising that it is the only opportunity to tell their story (Reference LeoLeo 1996b; Reference Leo and WhiteLeo and White 1999). They also may ask routine booking questions before they issue a warning, during which time they may establish rapport and predispose youths to waive (Pennsylvania v. Muniz 1990; Rhode Island v. Innis 1980; Reference WeisselbergWeisselberg 2008).

In about half of cases (52.8 percent), police gave the Miranda warning immediately after identifying the suspect. In the other half of cases (47.2 percent), police asked juveniles booking questions—name, age and date of birth, address and telephone number, grade in school, and the like—and sometimes used juveniles' responses to engage in casual conversations, to put youths at ease, and to accustom them to answering questions.

Police predispose youths to waive by emphasizing the importance to tell the truth, by nodding while reading the warning to cue the suspect to agree, or by telling the person that the interview constitutes his only opportunity to tell his story (Reference Redlich, Drizin, Kessler and KrausRedlich & Drizin 2007). Training manuals instruct police to blend the warning into the conversation, to describe it as a formality, or to summarize evidence, which a suspect can explain only if he waives (Reference WeisselbergWeisselberg 2008).

Police sometimes framed a Miranda waiver as a prerequisite to a juvenile's opportunity to tell his side of the story. Police communicated the value of talking—“telling her story”—and telling the truth before they gave a Miranda warning. Officers characterized the warning as an administrative formality to complete before the suspect can talk. Officers sometimes referred to the warning as “paper-work” to emphasize its bureaucratic quality or as a ritual with which to comply. A waiver form provides another opportunity to convert Miranda into a bureaucratic exercise.

Dickerson v. United States noted that Miranda warnings “have become part of our national culture” (Dickerson v. United States 2000: 430). Police invoke its cultural pervasiveness to minimize the warnings by “referring to their dissemination in popular American television shows and cinema, perhaps joking that the suspect is already well aware of his rights and probable can recite them from memory” (Reference Leo and WhiteLeo & White 1999:434–35). Officers regularly referred to youths' familiarity with Miranda from television and movies. Miranda's cultural ubiquity may detract from youths' understanding, as the warning becomes background noise at an interrogation.

After police warn a suspect, he or she must either waive or clearly invoke the rights to silence and to counsel (Berghuis v. Thompkins 2010; Davis v. United States 1994; Fare v. Michael C. 1979). Police establish that a juvenile understands his or her rights by reading the warning and then eliciting an affirmative response. In this study, officers read each right to the youth followed by the question “Do you understand that?” Juveniles acknowledged receiving each warning on-the-record—the Scales tape—and, in some departments, initialed and signed a Miranda form. Police in this study consistently obtained express waivers. After they ascertained juveniles understood the warning, they concluded the waiver process, “Bearing in mind that I'm a police officer and I've just read your rights, are you willing to talk to me about this matter?” Another version of the waiver formula ended, “Having these rights in mind, do you wish to talk to us now?”

Miranda reasoned that police must warn a suspect to dispel the inherent coercion of custodial interrogation. Justice White's Miranda dissent asked why those compulsive pressures do not coerce a waiver as readily as an unwarned statement (Miranda v. Arizona 1966). Legal analysts and criminologists concur that after police isolate a suspect in a police-dominated environment, a warning cannot adequately empower them to invoke their rights (Reference WeisselbergWeisselberg 2008; Reference WhiteWhite 1997). Post-Miranda studies consistently report about 80 percent or more of adults waive Miranda (Reference Cassell and HaymanCassell & Hayman 1996; Reference KassinKassin et al. 2007; Reference LeoLeo 1996b, Reference Leo2008; Reference WaldWald et al. 1967).

Juveniles waive Miranda rights at somewhat higher rates than do adults. Three decades of research reports that more than 90 percent of juveniles waive Miranda rights (Reference Goldstein and GoldsteinGoldstein & Goldstein 2010; Reference GrissoGrisso 1980; Reference Grisso and PomiciterGrisso & Pomiciter 1977). Juveniles' higher waiver rates may reflect their lack of understanding or inability to invoke Miranda effectively. Equally plausible, waivers may reflect prior justice system involvement and juveniles will have had less experience than adults (Reference Viljoen and RoeschViljoen & Roesch 2005). Table 2 reports that the vast majority of youths (92.8 percent) waived Miranda. This high rate is consistent with other juvenile studies and ten-percent higher than rates reported for adults. Interviews with justice system personnel confirmed the accuracy of these findings—almost all delinquents waived Miranda.Footnote ⁵

Table 2. Juveniles Who Waive or Invoke by Offense and Prior RecordFootnote *

^* Statistically Significant at: χ²(1, N = 300) = 5.7, P < 0.05.

^a Seven juveniles (2.3%) initially waived their Miranda rights and subsequently invoked them during interrogation, at which point interrogation ceased. Because they were truncated interrogation, I exclude them from analyses of police interrogation tactics.

Fare v. Michael C. (1979) cited his prior experience with police when it found a valid waiver. Analysts report a relationship between prior police contacts and Miranda invocations (Reference KassinKassin 2005; Reference LeoLeo 1996b). Post-Miranda research reported that defendants with prior arrests and felony convictions gave fewer confessions than did those with less experience. Older youths and those with prior felony referrals invoked more frequently than did younger juveniles and those without prior contacts (Reference GrissoGrisso 1980; Reference Grisso and PomiciterGrisso & Pomiciter 1977).

About one-third (35.1 percent) of these youths had one or more felony arrests prior to the offense for which police questioned them. Juveniles with one or more prior felony arrests waived their rights at significantly lower rates (86.9 percent) than did those with fewer or less serious police contacts (94.9 percent). Several factors likely contribute to more invocation by those with more extensive police contacts. Youths who waived at prior interrogations may have learned that confessing redounds to their disadvantage. The amount of time youths spend with lawyers contributes to greater understanding of rights, and those with prior arrests have more learning opportunities. Youths questioned previously may have learned to cope with and resist the pressures of interrogation.

Maximization and Minimization Interrogation Tactics

Police question suspects to obtain incriminating admissions or leads to other evidence—physical evidence, other participants, witnesses, or stolen property—which strengthen prosecutors' cases and facilitate guilty pleas. They seek suspects' statements—true or false—to pin them down, to control changes they later make in their stories, and to impeach their credibility. Police often described their roles to the two hundred eighty five juveniles who waived Miranda as dispassionate fact-finders. Minnesota training advises officers to portray themselves as neutral report writers who want to learn what happened to put in a statement for prosecutors and judges to evaluate (Reference NelsonNelson 2006). They frequently advise suspects that the interview provides their opportunity to “tell their story.” Depending on how forthcoming a youth is initially, they may use maximization and minimization tactics to elicit a statement.

Detectives may overstate a crime's seriousness, confront suspects with real or false evidence, accuse them of lying, challenge inconsistencies, emphasize the implausibility of their stories, and describe the negative impact that false statements would have on prosecutors and judges. The Reid Method instructs police to ask emotionally-charged Behavioral Analysis Interview (BAI) questions early in a suspect's interview to provoke a reaction (Reference InbauInbau et al. 2004). Although Reid advises officers to use BAI questions at a preliminary interview to screen the likely innocent from the probably guilty, few of these files indicated that police had any conversations prior to Scales recordings and none in which they interrogated these youths.

Police reported that they used maximization techniques regularly. They initially encouraged a suspect to commit to a story—true or false—and then used more confrontational tactics to challenge her version thereafter. Table 3 summarizes maximization strategies police used: confronted juveniles with evidence (54.4 percent); accused them of lying (32.6 percent); exhorted them to tell the truth (29.5 percent); asked BAI questions (28.8 percent); challenged inconsistencies (20.0 percent); emphasized the seriousness of the offense (14.4 percent); and accused them of other crimes (8.4 percent).

Table 3. Maximization Questions: Types and Frequency

In nearly one-third (30.9 percent) of interviews, police did not use any maximization techniques. In another quarter (23.1 percent) of interrogations, they used only one, which suggests that most juveniles did not require a lot of persuasion or intimidation to cooperate. Police used three or more maximization tactics in fewer than one-third (31.6 percent) of cases.

In about half (54.4 percent) the interrogations, police confronted juveniles with statements from witnesses or co-offenders, or referred to physical evidence. In most cases, DNA, surveillance, or fingerprint evidence will not be available in the short time between a suspect's arrest and interrogation. Sometimes, police described an investigation as if they already had obtained the evidence. In other instances, they questioned youths about potential evidence that later investigation would reveal. They asked a juvenile how he would respond to hypothetical evidence—“what if I told you” that someone had identified him or police found his fingerprints? In another version, officers might ask a juvenile “is there any reason why” his DNA might be on a gun or he would appear on surveillance video?

In about one-third (32.6 percent) of cases, officers accused juveniles of lying. Police typically allowed juveniles to commit to a story and then confronted them. In nearly one-third of cases (29.5 percent), officers urged juveniles to be honest and tell the truth.

Officers reaffirmed their roles as objective fact-gatherers and neutral conduits who would accurately convey juveniles' statements to prosecutors and judges. Police intimated that their recommendations could affect prosecutors' charge evaluations and judges' decisions. They cautioned that prosecutors and judges reacted negatively to an implausible story and predicted that judges responded more favorably to truthful defendants.

Inbau and Reid advise interrogators at a preliminary interview to ask “behavior-provoking questions that are specifically designed to evoke behavioral responses” (2004:173). They posit that innocent and guilty people respond differently to emotionally provocative questions and enable investigators accurately to classify them. Leo (Reference Leo1996b) reported that officers asked BAI questions in about 40 percent of interrogations. In this study, police used BAI questions in more than one-quarter (28.8 percent) of interviews, most commonly “Do you know why I have asked to talk to you here today?”

In one-fifth (20 percent) of cases, officers confronted and challenged suspects' assertions. They pointed out inconsistencies, disputed claims, and questioned youths' credibility to increase anxiety and undermined confidence. Officers regularly responded to juveniles' claims of innocence with a barnyard epithet—“Bullshit.”

Many interviews began with an invitation to a youth to tell his story. But, police warned that it was a time-limited opportunity. Officers cautioned that if youth did not take advantage of this chance to explain their involvement, then they might regret it later. Police withheld information from juveniles about the investigation to increase uncertainty and anxiety and cautioned a reluctant youth that without his version, other co-offenders might shift responsibility to him or make a deal at his expense.

Although police contamination—disclosure of information known only to police or a true perpetrator about the crime and later incorporated by a suspect—is a recurring theme in studies of false confessions (Reference GarrettGarrett 2010, 2011; Reference LeoLeo 2008), contamination appears unlikely in these interrogations. As will be seen, interrogations were surprisingly brief, most youths confessed or made admissions at the outset, and officers confronted youths with any evidence in only half (54.4 percent) the cases.

Minimization tactics offer face-saving excuses or moral justifications that reduce a crime's seriousness, provide a less odious motivation, or shift blame to a victim or accomplice (Reference KassinKassin et al. 2010). Themes imply a suspect will feel better or will derive benefit if he confesses (Reference LeoLeo 2008). Table 4 reports that police used minimization tactics in fewer than one-fifth (17.3 percent) of these interrogations, far less often than they used any maximization tactics (69.1 percent). Although prosecutors charged all these youths with felonies, one officer explained that “most of these are fairly minor, so you don't have to do a whole lot of minimizing.” Officers used scenarios or themes to reduce suspects' guilt or culpability in 15.4 percent of cases; appealed to self-interest in one-tenth (11.9 percent) of cases; expressed empathy in one-tenth of cases (10.5 percent); and used other tactics in a few cases. The relative paucity of minimization tactics is consistent with research in the United Kingdom, and Minnesota training that discourages their use (Reference SoukaraSoukara et al. 2009; Reference NelsonNelson 2006).

Table 4. Minimization Questions: Types and Frequencies

The Reid Method teaches police to develop a theme or scenario to neutralize guilt, minimize responsibility, and make it easier to confess (Reference InbauInbau et al. 2004). Criminologists have used techniques of neutralization to understand how youths rationalize delinquent behavior (Reference Sykes and MatzaSykes & Matza 1957). Many themes are extensions of criminal law defenses—provocation, intoxication, or insanity—that provide rationales to reduce moral constraints (Reference MatzaMatza 1964). For example, delinquents may reject mental illness—insanity—as an excuse, but embrace the idea of “going crazy” or “being mad” to rationalize criminal conduct. Police sometimes suggested that getting mad, losing control, or excitement accounted for youths' criminal misconduct. Intoxication provides an explanation for bad behavior, and juveniles readily invoked drinking alcohol or using drugs to excuse criminal conduct.

Police diffused juveniles' responsibility by suggesting that they succumbed to negative peer influences. Juveniles often commit their crimes in groups (Reference Snyder and SickmundSnyder & Sickmund 2006), and police can blame others and allow juveniles to shift blame as well. Parents regularly refer to errant children's behavior as a mistake and youths learn that mistakes can mitigate responsibility. Police regularly encouraged juveniles to attribute their delinquency to a mistake.

Police described benefits juveniles might derive and appealed to self-interest in one-tenth (11.9 percent) of cases. They offered to investigate further and assist juveniles to receive help. They intimated that prosecutors and judges would view more favorably youths who confessed than those who lied and might deal with them more leniently.

Officers minimized seriousness by describing the triviality of a youth's crime compared with the gravity of other delinquents' offenses. Even a serious crime—a drive-by shooting—could have been worse if the shooter had hit the intended target. The rationale of juvenile courts—treatment rather than punishment—provided officers with another theme with which to offer help and to minimize seriousness.

Juveniles' Responses

I examined how the 285 youths who waived Miranda responded to police and how their attitudes affected how much information they provided. I classified the outcome of an interrogation based on the evidentiary value of a statement (Reference Cassell and HaymanCassell & Hayman 1996; Reference WaldWald et al. 1967). Table 5 reports outcomes of interrogations which are coded into three categories—confess, admit, or deny.Footnote ⁶

Table 5. Outcome of Interrogation and Youths' Attitude

^* Statistically Significant at: χ²(1, N = 285) = 7.84, P < 0.001.

A majority (58.6 percent) of juveniles confessed within a few minutes of waiving Miranda and did not require prompting by police. British research confirms that the majority of suspects confessed and “almost all did so near the beginning of the interviews” (Soukara et al. Reference Soukara2009:495). UK analysts conclude that “suspects enter a police interview having already decided whether to admit or deny the allegations against them” and interrogation tactics have little impact on whether they admit (Milne & Bull Reference Milne and Bull1999:81).

An additional one-third (29.8 percent) of juveniles provided statements of some evidentiary value, for example, admitting that they served as a look-out during a robbery or participated in a burglary even if they did not personally steal property. Justice personnel agreed that most juveniles made some incrimination admissions.

Other studies corroborate similar high rates of admissions and confessions. Leo (Reference Leo1996b) found such outcomes in three-quarters (76 percent) of cases in which adults waived Miranda. The Yale-New Haven study reported that about two-thirds (64 percent) of interrogations produced incriminating evidence (Wald et al. Reference Wald1967). A survey of police investigators estimated that two-thirds (68 percent) of suspects made incriminating statements (Kassin et al. Reference Kassin2007). Other UK research reports a rate of 77 percent, ranging from 64 percent to 97 percent among various police stations (Bull & Milne Reference Bull, Milne and Glaser2004; Evans Reference Evans1993). More than half (55 percent) of delinquents held in detention reported they had confessed (Viljoen et al. Reference Viljoen2005).

Only a small proportion (11.6 percent) of juveniles made no incriminating admissions. Without invoking Miranda outright, forms of resistance included non-cooperation, denial of knowledge and culpability, lying, evasion, silence, or blame shifting. When confronted with resistance, police used more maximization techniques than they did with cooperative youths, but did not question them for longer periods. Once they recognized a youth was resistant, they concluded the interview with the observation that prosecutors and judges who reviewed their interrogation would not view them favorably.

Criminologists have studied the interplay between police discretion and juveniles' attitudes (Clarke & Sykes Reference Clarke, Sykes and Glaser1974). For less serious crimes, deferential youths reduce likelihood of arrest and contumacious ones increase it (LaFave Reference LaFave1965; Skolnick Reference Skolnick1967; Pilliavin & Briar Reference Pilliavin and Briar1964; Bittner Reference Bittner and Rosenheim1976). Studies of police and probation officers report that a youth's attitude affected how officials perceived, imputed moral character, and responded to them (Cicourel Reference Cicourel1995; Emerson Reference Emerson and Glaser1974). When youths' attitudes affect police decisions, minority youths typically fare worse than do their white counterparts (Bittner Reference Bittner and Rosenheim1976; Black & Reiss Reference Black and Reiss1970).

Police reported that juveniles' attitudes ranged the gamut—“some are scared to death, and others, it's almost a joke.” Many officers described youths as scared, especially “the kids that are new to the process.” Although police described some youths as confrontational, justice system personnel viewed most youths as compliant or submissive. “I would say that 90 percent or more would probably be cooperative and the other percentage would be the frequent-fliers so to speak.” Several officers used the same expression—“deer in the headlights”—to describe youths' demeanor in the interrogation room. Public defenders described most juveniles as humbled or defeated when they confessed.

Ethnographers emphasize the importance of attitude—“rude or impolite, aggressive or passive, laughter or tears, and the like”—and its impact on justice system processing (Cicourel Reference Cicourel1995:xv). Juveniles exhibited many attitudes during interrogation—polite, cooperative, distressed, remorseful, frightened, cocky, resistant, aggressive, and confrontational—which could fluctuate from one minute to the next.

Police reports frequently included comments about juveniles' demeanor and behavior during interrogation. They documented whether they believed suspects told the truth or lied and indicated whether they cooperated or resisted. Officers often described youths' emotional or behavioral responses to their interrogators. Based on my impressions and officers' reports, I dichotomized attitude as cooperative or resistant. Other research used similar categories and described eighty percent of suspects as cooperative (Baldwin Reference Baldwin1993). Juveniles cooperate for many reasons—human decency in social interactions, fear and anxiety, dependency on authority figures, or the coercive pressures of isolation—but most exhibited positive attitudes.

As Table 5 indicates, the vast majority of juveniles (79.6 percent) exhibited a cooperative demeanor and only one-fifth (20.4 percent) appeared resistant. Not surprisingly, the vast majority (96.5 percent) of cooperative juveniles confessed or made incriminating admissions. By contrast, fewer than one-in-ten (8.5 percent) resistant juveniles confessed and almost half (43.1 percent) provided no useful admissions. Only one-tenth (11.6 percent) of youths denied involvement, but those who exhibited resistant attitudes accounted for more than three-quarters of them (75.8 percent).

Police question suspects to elicit admissions or obtain statements that prosecutors can use to impeach testimony. Suspects' answers may lead to other evidence—witnesses, co-defendants, or property. Table 5 reports the proportion of cases in which interrogations yielded corroborating evidence. I defined corroborating evidence as evidence which police did not possess prior to questioning—leads to physical evidence, a crime scene diagram, identity of a co-offender, or unknown witness. By this conservative standard, fewer than one-fifth (18.2 percent) of interviews yielded information that police did not already have. Interrogation did not produce much collateral evidence and gathering it appears to be a secondary goal.

Some police attributed the relatively low-yield of corroborating evidence to time pressure and volume of cases under which they labored. Once police obtained an admission—which they did quickly—they did not press youths for additional evidence. Prosecutors confirmed that interrogations did not often lead to corroborating evidence, but they attributed that to good preliminary investigations. Police questioned more than two-thirds (69.7 percent) of juveniles within less than 24-hours of their crimes—effectively, they were “caught in the act.” Police and prosecutors had strong enough evidence with which to convict youths without an interview in about two-thirds (63.2 percent) of cases. Police and prosecutors said that juveniles' statements often provided bases to obtain search warrants, which produced additional evidence not disclosed by the interview.

Length of Interrogation

It can take a long time and rigorous questioning to elicit a false confession. Although police may obtain some false confessions within an hour or two, they elicited eighty-five percent of false confessions after suspects had been in custody or interrogated for six hours or longer (Drizin & Leo Reference Drizin and Leo2004). Table 6 reports the length of interrogations, length of time by type of offense, and length of time by whether the offense involved a firearm.Footnote ⁷ Routine felony interrogations are brief. Police completed three-quarters (77.2 percent) of interviews in less than fifteen minutes and concluded nine-in-ten (90.5 percent) in less than thirty minutes. In the longest interviews, police questioned three youths (1.1 percent) for more than one and one-half hours. Although prosecutors charged youths with one or more felonies, brief interrogations are unlikely to cause false confessions (White Reference White2001).

Table 6. Length of Interrogation by Type of OffenseFootnote * and WeaponFootnote **

^* Statistically Significant at: χ²(1, N = 285) = 32.3, P < 0.05.

^** Statistically Significant at: χ²(1, N = 285) = 9.4, P < 0.01.

Although these short interviews initially seemed surprising, other research confirms that interrogations of even two or three hours are exceptional and frequently problematic (Drizin & Leo Reference Drizin and Leo2004; Kassin et al. Reference Kassin2007). The Yale-New Haven study reported that police questioned suspects for more than an hour in only 15 percent of cases (Wald et al. Reference Wald1967). Leo (Reference Leo1996c) reported that police questioned only one-quarter (28.7 percent) of suspects for more than one hour. Cassell and Hayman (Reference Cassell and Hayman1996) reported that only 13 percent of interrogations took more than 30 minutes and only one lasted longer than an hour. Research on British interrogations of juveniles reported that “[i]nterviews tended to be very brief with the majority taking less than fifteen minutes (71.4 percent). Although the average length of interviews was around 14 minutes, the most frequent length was around 7 minutes” (Evans Reference Evans1993:26). Analyses of taped UK interrogations reported that “most were short and surprisingly amiable discussions” in which more than one-third of suspects confessed at the outset (Baldwin Reference Baldwin1993:331). Kassin and Gudjonsson (Reference Kassin and Gudjonsson2004:46) summarized research and reported that “[m]ost of the interviews were short (80 percent lasted less than 30 minutes; 95 percent were completed within 1 hour), the confession rate was 58 percent, little interrogative pressure was applied, and very few suspects who initially denied guilt eventually confessed.” Inbau et al. (Reference Inbau2004) warn against interrogations that last longer than four hours, a duration substantially longer than observed in any research. By contrast, police extracted most false confessions only after interrogations of six hours or longer (Drizin & Leo Reference Drizin and Leo2004).

I asked justice professionals to estimate the lengths of interviews and they universally agreed, “They're actually very short.”Footnote ⁸ When asked why police concluded felony interrogations so quickly, justice system personnel attributed brevity to several factors. Many professionals referred to police workload pressures. Police conducted a form of triage and questioned suspects longer in more serious cases, but did not regard most juvenile felonies as serious crimes. Several officers attributed brief interrogations to the relative simplicity of most youth crime and their ability to elicit admissions quickly.

In this study, a statistically significant relationship appeared between length of interrogation and type of offense. Police questioned more youths charged with property and drug crimes for fifteen minutes or less than they did youths charged with other types of offenses. Crimes that involved some physical evidence—drugs, stolen property, or automobiles—may have provided police with more evidence with which to confront these juveniles.

Cases involving firearms resulted in longer interrogations. Although police questioned only 9.5 percent of suspects for longer than thirty minutes, they interrogated twice as many (20 percent) juveniles charged with firearms offenses for longer than thirty minutes. I compared the lengths of interrogation in all cases that involved guns—armed robbery, assault with a gun, firearms possession, or burglary in which youths stole guns—with cases in which juveniles used other weapons—knives, blunt instruments, or automobiles—or did not use a weapon.

Guns provide an indicator of offense seriousness (Reference Podkopacz and FeldPodkopacz & Feld 1996), and police questioned these juveniles longer and more aggressively. Guns affected the tactics as well as the length of interrogations. Police wanted to recover guns used or stolen by youths, and they used maximization and minimization tactics more extensively to retrieve them. Officers referred to the benefits that would accrue to a youth who helped to recover a gun and described the dangers guns posed to people who held them and those around them. Only two interrogations in this study raised constitutional issues of voluntariness and both involved questioning to recover guns. In each case, police questioned juveniles for the longest time (one and one-half to two hours), used the most maximization techniques, and made explicit quid-pro-quo promises of leniency to recover guns used or stolen by juveniles.

Police and justice system personnel confirmed the relationship between guns and length of interrogation and agreed that guns provide a proxy for seriousness. Police associated guns with youths' involvement with gangs—another indicator of seriousness. Police questioned youths to recover the gun and learn about other youths who had contact with the weapon. Youths knew that gun crimes garnered serious consequences, raised the stakes, and gave them greater incentive to resist interrogators. Serious crimes are more likely to go to trial and police invested more energy to strengthen prosecutors' cases.

Protecting Young Offenders in the Interrogation Room

Although Miranda purported to bolster the adversary system and protect citizens, warnings failed to achieve those goals. Decisions since Miranda have limited its scope and applicability and adverse consequences when police fail to comply (Reference SloboginSlobogin 2007; Reference WeisselbergWeisselberg 2008). Miranda's assumption that a warning would enable suspects to resist the compulsive pressures of interrogation is demonstrably wrong. Post-Miranda research reports that the vast majority of suspects waive and only some sophisticated suspects invoke. Eighty percent of adults and ninety percent of juveniles waive their sole protection in the interrogation room. Although Miranda recognized that those compulsive pressures threatened the adversarial process, waivers provide police with a window of opportunity to conduct an inquisitorial examination. Perversely, Miranda allows judges to focus on ritualistic compliance with a procedural formality rather than to assess the voluntariness or reliability of a statement (Reference GodseyGodsey 2005; Reference LeoLeo 2008; Reference WeisselbergWeisselberg 2008). Judicial review of a Miranda waiver is the beginning and end of regulating interrogation (Missouri v. Seibert 2004).

Miranda is especially problematic for younger juveniles who may not understand its words or concepts. The Court has recognized youths' vulnerability and distinguished between younger and older youths. Developmental psychologists corroborate their differing abilities. Younger juveniles' incomplete understanding, impaired judgment, and heightened vulnerability warrant greater assistance—a non-waivable right to counsel—to assure voluntariness of a Miranda waiver and statement. Psychologists distinguish between youths' cognitive ability—capacity to understand—and ability to make mature decisions and exercise self-control. Miranda requires only the ability to understand words, which developmental psychologists conclude that most sixteen-and seventeen-year-old youths can do.

This study corroborates that sixteen- and seventeen-year-old juveniles appear to understand and exercise Miranda similarly to adults. This consistency inferentially bolsters research that younger juveniles increasingly lack ability to understand and competence to exercise rights. Psychologists report that many, if not most, children fifteen or younger do not understand Miranda or possess competence to make legal decisions (Reference GrissoGrisso 1980; Reference GrissoGrisso et al. 2003). Research on false confessions underscores the unique vulnerability of younger juveniles (Reference Drizin and LeoDrizin & Leo 2004; Garrett 2011; Reference GrossGross et al. 2005). Police obtained more than one-third (35 percent) of proven false confessions from suspects younger than eighteen (Reference Drizin and LeoDrizin & Leo 2004), and younger adolescents are at greater risk to confess falsely than older ones (Reference TepferTepfer et al. 2010).

Developmental psychologists attribute their over-representation among false confessors to reduced cognitive ability, developmental immaturity, and increased susceptibility to manipulation (Reference Bonnie, Grisso, Grisso and SchwartzBonnie & Grisso 2000; Reference Tobey, Grisso and SchwartzTobey et al. 2000; Reference RedlichRedlich 2004). They have fewer life experiences or psychological resources with which to resist the pressures of interrogation (Reference Drizin and LuloffDrizin & Luloff 2007; Reference RedlichRedlich 2004). Juveniles' lower social status and societal expectations of obedience to authority create pressures to waive (Reference GudjonssonGudjonsson 2003; Reference LeoLeo 2009). Juveniles are more likely than are adults to comply with authority figures, tell police what they think they want to hear, and respond to negative feedback (Reference GudjonssonGudjonsson 2003; Reference LyonLyon 1999). The stress and anxiety of interrogation intensify their desire to extricate themselves in the short-run by waiving and confessing (Reference Goldstein and GoldsteinGoldstein & Goldstein 2010; Reference Owen-KostelnikOwen-Kostelnik et al. 2006). Impulsive decision-making and limited ability to consider long-term consequences heighten their risk (Reference Redlich, Lassiter and MeissnerRedlich 2010). The immature adolescent brain contributes to impulsive behavior and heightened vulnerability (Reference BirckheadBirckhead 2008; Reference MaroneyMaroney 2009).

Despite youths' heightened susceptibility, police do not incorporate developmental differences into the tactics they employ (Reference Owen-KostelnikOwen-Kostelnik et al. 2006). Techniques designed to manipulate adults—aggressive questioning, presenting false evidence, and leading questions—may create unique dangers when employed with youths (Reference Kaban and TobeyKaban & Tobey 1999; Reference Redlich, Drizin, Kessler and KrausRedlich & Drizin 2007; Reference Tanenhaus and DrizinTanenhaus & Drizin 2002). Police in this study did not report receiving special training to question juveniles and used the same tactics that Leo (Reference Leo1996b) reported they employed with adults.

The Court in Haley, Gallegos, Gault, Fare, Alvarado, and J.D.B. excluded statements taken from youths fifteen years of age or younger and admitted those obtained from sixteen- and seventeen-year-olds. The Court's de facto line—fifteen and younger versus sixteen and older—closely tracks what psychologists have found about youths' ability to understand the warning and concepts. State courts and legislatures should formally adopt the functional line that the Court and psychologists discern between youths sixteen and older and those fifteen and younger.

Analysts advocate that juveniles younger than sixteen years of age “should be accompanied and advised by a professional advocate, preferably an attorney, trained to serve in this role” (Reference KassinKassin et al. 2010:28). Juveniles should consult with an attorney, rather than to rely on parents, before they exercise or waive constitutional rights (Reference Bishop and FarberBishop & Farber 2007; Reference FarberFarber 2004). More than three decades ago, the American Bar Association endorsed mandatory, non-waivable counsel because it recognized that “Few juveniles have the experience and understanding to decide meaningfully that the assistance of counsel would not be helpful” (American Bar Association, Institute of Judicial Administration 1980:92).

Requiring a child to consult an attorney assures an informed and voluntary waiver (Reference Drizin and LuloffDrizin & Luloff 2007; Reference FarberFarber 2004). If youths fifteen years of age or younger consult with counsel prior to waiver, it will limit somewhat police's ability to secure confessions. However, if most younger juveniles cannot understand and exercise rights without legal assistance, then to treat them as if they do denies fundamental fairness and enables the state to exploit their vulnerability. Constitutional rights exist to assure factual accuracy, promote equality, and protect individuals from governmental over-reaching, and inevitably diminish the state's ability to fight crime (Escobedo v. Illinois 1964). Fare emphasized lawyers' unique role in the justice system, and Haley, Gallegos, and Gault recognized younger juveniles' exceptional need for assistance.

Limiting the Length of Interrogations

Most false confessions emerge only after lengthy interrogations and youthfulness exacerbates those dangers (Reference Drizin and LeoDrizin & Leo 2004; Reference GrossGross et al. 2005; Reference Kassin and GudjonssonKassin & Gudjonsson 2004). The Court recognizes that lengthy interrogations can produce involuntary confessions (Ashcraft v. Tennessee 1944), and found that questioning juveniles for even five or six hours could produce an involuntary statement (Gallegos v. Colorado 1962; Haley v. Ohio 1948). Policy-makers should create a sliding-scale presumption of involuntariness based on length of interrogation and examine a confession's reliability more closely as length of questioning increases (Reference LeoLeo et al. 2006).

The vast majority of interrogations are very brief. Police concluded ninety percent of these felony interrogations in less than thirty minutes. Every study reviewed reports that police completed nearly all interrogations in less than an hour and few take as long as two hours. By contrast, interrogations that elicit false confessions are usually long inquiries that wear-down an innocent person's resistance—eighty-five percent took at least six hours (Reference Drizin and LeoDrizin & Leo 2004).

This study does not enable me to prescribe outer time limits because I did not encounter either lengthy or factually problematic interrogations. However, states should create a sliding-scale presumption that police elicited an involuntary confession as the length of questioning increases. Police complete nearly all felony interrogations in less than one hour, but extract most false confessions only after grilling suspects for six hours or longer. These times provide a framework to limit interrogations and strengthen the presumption of coercion. The contrast between the short duration of routine questioning and lengthy interrogation associated with false confessions leads analysts to conclude that interrogation should not extend more than four hours (Reference KassinKassin et al. 2010; Reference WhiteWhite 2001). Four hours provides ample opportunity for police to obtain true confessions from guilty suspects willing to talk without coercion without increasing the risk of eliciting false confessions from innocent people.

On the Record

Within the past decade, legal scholars, psychologists, law enforcement, and justice system personnel have reached consensus that recording interrogations reduces coercion, diminishes dangers of false confessions, and increases reliability (Reference CassellCassell 1998; Reference Drizin and ReichDrizin & Reich 2004; Reference GarrettGarrett 2010, 2011; Reference GudjonssonGudjonsson 2003; Reference Milne and BullMilne & Bull 1999; Reference SullivanSullivan 2004, Reference Sullivan2006, Reference Sullivan, Lassiter and Meissner2010). About a dozen states require police to record interrogations, albeit some under limited circumstances—homicide or young suspects (Garrett 2011; Reference LeoLeo 2008; Reference Sullivan, Lassiter and MeissnerSullivan 2010). Many police departments have policies to record interrogations for some crimes (Reference SullivanSullivan 2006, Reference Sullivan, Lassiter and Meissner2010).

Recording creates an objective record and an independent basis to resolve disputes between police and defendants about Miranda warnings, waivers, or statements (Reference SloboginSlobogin 2003). A complete record enables fact-finders to decide whether a statement contains facts known to a guilty perpetrator or police supplied them to an innocent suspect during questioning (Reference GarrettGarrett 2010, 2011; Reference WhiteWhite 1997). Recording protects police from false claims of abuse (Reference CassellCassell 1998; Reference WhiteWhite 1997). It enables police to focus on suspects' responses, to review details of an interview not captured in written notes, and to test them against subsequently discovered facts (Reference Drizin and ReichDrizin & Reich 2004). It reduces the need for an officer to take notes or a second person to witness a statement, which may chill a suspect's willingness to talk. Recording avoids distortions that occur when interviewers rely on memory or notes to summarize a statement (Reference Milne and BullMilne & Bull 1999).

A recorded confession greatly strengthens prosecutors' plea bargain advantage. It enables them to avoid suppression hearings, negotiate better pleas, and obtain convictions (Reference SullivanSullivan 2006; Reference WhiteWhite 1997). Defense lawyers can review recordings rather than rely on clients' imperfect recollection of a stressful event. Scales recordings have virtually eliminated motions to suppress confessions because tapes provide unimpeachable evidence. This generates substantial savings because police, prosecutors, and defense counsel do not have to prepare for and judges do not have to conduct suppression hearings about Miranda warnings, coercive tactics, or the accuracy of a statement.

Police must record all conversations—preliminary interviews and interrogations—rather than just a final statement—a “post-admission narrative” (Reference GudjonssonGarrett 2011; Gudjonsson 2003). Otherwise, police may conduct a pre-interrogation interview, elicit incriminating information, and then record only a final confession after the “cat is out of the bag”—a variation of the two-step practice condemned in Missouri v. Siebert. Only a complete record of every interaction can protect against a final statement that ratifies an earlier coerced one or against a false confession contaminated by non-public facts that police supplied a suspect (Garrett 2011; Reference KassinKassin 1997).

The Court repeatedly insists that American criminal and juvenile justice is an adversary system. Repeated assertions do not alter the reality that states establish most defendants' guilt through an inquisitorial system in which suspects seal their fate in the interrogation room and render trial procedures a nullity—interrogation elicits confessions and confessions produce guilty pleas. Because states do not and need not provide full adversarial testing in every case, we need stronger mechanisms to assure factual reliability of inquisitorial justice. Recording imposes no great burden on police, illuminates the inner-workings of the interrogation room, and provides an objective record on which a defendant may appeal to a judge. Because the vast majority of defendants do not receive a trial, judicial review of the record provides an alternative check to assure the reliability of routine felony justice.

Data Limitations and More Research

This study suffers from several methodological limitations. There is a major problem of sample selection bias. I could not randomly select the files I analyzed from a larger universe of interrogations because such an array of cases simply does not exist. The study includes only juveniles whom prosecutors actually charged with felonies and whose files contained a record of their invocation or interrogation. Because of confidentiality restrictions on data access, I personally coded all of the interrogations and could not use multiple coders or obtain inter-rater reliability scores. The four counties in this study represent about half the population and delinquency petitions in Minnesota. Because Scales had required Minnesota police to record interrogations for more than a decade, police practices probably differ from and may be more benign that those used in other states or for more serious crimes. Despite these caveats, this study represents the largest number of interrogations and of juveniles in the criminological literature. While not necessarily a representative sample, it provides important insights into routine police questioning of older juveniles and a baseline of practice against which to evaluate conditions of interrogation likely to elicit false confessions.

This study is only the second naturalistic empirical study of police interrogation in the United States in the past three decades (Reference LeoLeo 1996b), and the first involving juveniles. We need far more empirical research on interrogations practices in general, in a number of different settings, and with more knowledge about characteristics of suspects. As more jurisdictions adopt taping and recording requirements, we will have further opportunity to conduct this type of research. Recordings provide opportunities for psychologists, criminologists, police, and others to study systematically what actually occurs in the interrogation room. This will increase our fund of knowledge, enable us to develop more effective techniques to elicit true confessions from guilty defendants, reduce the likelihood of extracting false confessions from innocent suspects, and provide a stronger basis for systemic policy prescriptions.