Cumulative Disadvantage: A Psychological Framework for Understanding How Innocence Can Lead to Confession,Wrongful Conviction,and Beyond

Miranda—the right to remain silent and have legal representation

In the landmark case of Miranda v. Arizona (1966), the U.S. Supreme Court ruled that police must inform in-custody suspects of their constitutional rights to remain silent and have a lawyer present. Aimed at protecting citizens from the “inherently compelling pressures” of police interrogation, the Court offered a remedy: When suspects are in custody, police must inform them of these rights; any statement taken without a knowing, intelligent, and voluntary waiver would then be considered involuntary and not admissible into evidence.

Despite the Court’s objectives, the promise of Miranda has not been fulfilled. Subsequent court rulings have narrowed the protections to such an extent that legal scholars no longer see it as protective (e.g., Weisselberg, 2017; White, 2001). Meanwhile, research psychologists using standardized comprehension instruments have found that some people—notably young adolescents and cognitively impaired adults—do not comprehend the warnings they were given (e.g., Grisso, 1981; for a review, see Zelle, Romaine, & Goldstein, 2015; for findings mirrored by eye-tracking methodologies, see Scherr, Agauas, & Ashby, 2016); that the content, format, and language of these warnings vary across jurisdictions, resulting in disparities in difficulty level (Rogers, Harrison, Shuman, Sewell, & Hazelwood, 2007; Rogers, Hazelwood, Harrison, Sewell, & Shuman, 2008); and that the situational stress elicited by accusation can further impair comprehension (Rogers, Gillard, Wooley, & Fiduccia, 2011; Scherr & Madon, 2012). In short, empirical research has cast serious doubt on the protective adequacy of Miranda (for reviews, see Kassin, Scherr, & Alceste, 2019; Smalarz et al., 2016).

Relevant legal precedent

The original Miranda ruling noted that suspects must voluntarily waive their rights and this decision should not result from threat, trickery, or cajoling (Miranda v. Arizona, 1966). Even though such legal precedent exists, interrogators still seek to manipulate suspects at this stage and subsequent legal precedents have permitted the use of such tactics to obtain waivers (e.g., Clark v. Murphy, 2003; Hairston v. United States, 2006; United States v. Washington, 2006). Three rulings highlight the sheer degree to which recent precedent has veered from the original Miranda ruling and help to elucidate the relevant scientific research and the cumulative-disadvantage framework.

In Florida v. Powell (2010), the Court ruled that it is not necessary to clearly inform suspects that they can invoke their rights at any point during the interrogation. The danger of this precedent for innocents stems from their initial naivete motivating their decision to talk to interrogators to prove their innocence. In this scenario, innocents may eventually come to realize during the ensuing interrogation that they are helpless to convince the interrogator of their innocence. Yet without knowing that they can invoke their rights at any point, even after the initial waiver, they may go on to believe that confessing is their only means of escape.

In Berghuis v. Thompkins (2010), the Supreme Court puzzlingly ruled that suspects need to explicitly invoke their right to silence, essentially indicating that suspects must speak up to remain silent. This decision provides legal backing for protocols that unreasonably require suspects under stress to precisely articulate their wishes (e.g., “I wish to remain silent”). Not realizing that an explicitly precise articulation is necessary, innocent suspects may be especially likely to continue to talk in attempts to invoke their rights, but to no avail.

Although the Miranda court ruled that invoking one’s rights could not be used against them in court (Miranda v. Arizona, 1966; Skinnider & Gordon, 2001), a third Supreme Court decision suggested that such instances can be used against them in court (Salinas v. Texas, 2013). The Salinas ruling represents the most problematic divergence from the original Miranda ruling for all suspects—guilty and innocent alike. Innocents who waive their rights will face a professional highly confrontational interrogation without an attorney present. Conversely, those who exercise their right will face the potential that this decision could trigger an adverse inference. Thus, recent legal precedents have diminished the protection initially afforded by Miranda.

To summarize, the first stage of the cumulative-disadvantage framework is set into motion with an information-gathering interview in which behavioral cues, wholly lacking scientific support, are used to determine deception. As a result, interviews serve as a pivotal point to determine whether a suspect should be questioned further, in custody, and in an accusatory manner. A second impetus—a naive mind-set—motivates innocents, unlike their guilty counterparts, to make decisions and exhibit behaviors consistent with the ideas that they have nothing to hide or fear. As a result, the cumulative disadvantage process is triggered by investigators’ presumptions of guilt and a phenomenology of innocence that puts innocents at risk to waive their interrogation rights, which, consequently, sets these suspects on trajectories to navigate custodial interrogations and confront police intimidation in isolation.

The process of interrogation

“Third-degree” interrogation methods involving physical torture declined precipitously from the 1930s to the 1960s and were replaced by an approach that was more psychological in nature. One aim of psychological approaches to interrogation is to lessen the stress associated with confession relative to denial, making it easier for a suspect to confess as a means of escaping an unpleasant situation (for a general overview of the confession process, see Yang, Guyll, & Madon, 2017). Decades of evidence show that people prefer immediate outcomes over distal ones (Herrnstein, 1997; Navarick, 1982; Rachlin, Brown, & Cross, 2000), and adults are substantially more likely to admit to wrongdoing in order to avoid proximal consequences (Madon, Guyll, Scherr, Greathouse, & Wells, 2012; Madon, Yang, Smalarz, Guyll, & Scherr, 2013; Scherr, Miller, & Kassin, 2014; Yang, Madon, & Guyll, 2015). Myopic decision-making tendencies that impulsively satisfy an immediate concern despite some later consequence are especially likely among juveniles and substance users (e.g., Baker, Johnson, & Bickel, 2003; Bickel & Marsch, 2001; Bickel, Odum, & Madden, 1999; Kollins, 2003; Reynolds, Richards, Horn, & Karraker, 2004).

In an attempt to capitalize on suspects’ myopic tendencies, investigators are trained to create specific dynamics. Most popularly articulated in Inbau and Reid’s (1962) Criminal Interrogations and Confessions (for the most recent edition, see Inbau et al., 2013), the nine-step Reid technique advises isolating the suspect in a small, bare room to create a “nonsupportive” environment. Investigators achieve this goal, in part, as a consequence of innocents’ naive decision making highlighted in Stage 1. In atypical cases in which innocents exercise their rights, they are likely aided by legal counsel during the custodial interrogation, thereby minimizing the time they spend in social isolation and fundamentally altering the environment from nonsupportive to supportive.

Alongside the social isolation, interrogators confront suspects with combinations of negative and positive incentives, dubbed “maximization” and “minimization” (Kassin & McNall, 1991). On the one hand, the interrogator confronts the suspect with accusations of guilt, assertions that may be bolstered by evidence, real or false, and refuses to accept objections and denials. On the other hand, the interrogator offers sympathy and moral justification, introducing “themes” that minimize the crime and lead suspects to see confession as an expedient means of escape. The prevalent use of these techniques has been well documented (e.g., Cleary & Vidal, 2016; Feld, 2013b; Kassin et al., 2007; Kelly, Miller, Redlich, & Kleinman, 2013; King & Snook, 2009; Leo, 1996a; for critiques, see Kassin, 1997; for a historical account, see Leo, 2008).

Inbau et al. (2013) claim that their tactics produce outcomes that are “diagnostic,” that they elicit confessions from offenders but not from innocent suspects. Yet controlled research and a growing number of wrongful convictions indicate that these techniques do in fact put innocent suspects at risk to confess for two reasons. First, some suspects are dispositionally vulnerable to influence under pressure (Gudjonsson, 2003, 2018). Archival analyses, laboratory experiments, and self-report surveys conducted in the United States and throughout Europe have shown that juveniles are particularly vulnerable as suspects (Drizin & Leo, 2004; Goldstein, Condie, Kalbeitzer, Osman, & Geier, 2003; Gudjonsson, Sigurdsson, & Sigfusdottir, 2009; Malloy, Shulman, & Cauffman, 2014; Owen-Kostelnik et al., 2006; Redlich & Goodman, 2003). Second, research shows that certain psychological tactics that are commonly used and lawful in the United States can induce false confessions, even from adults who are relatively well adjusted and competent. Two tactics in particular pose a risk: the presentation of false evidence and minimization themes that communicate leniency.

Relevant legal precedent

The fact that interrogators in the United States are permitted to lie to suspects and strongly imply leniency astonishes most people. The evolution to using psychology-based interrogation approaches was likely an unintended result of Brown v. Mississippi (1936), wherein the U.S. Supreme Court banned the use of physical third-degree tactics (Leo, 2008). Yet despite the recognition that confessions often render other evidence unnecessary (Colorado v. Connelly, 1986) and the apt realization of the need to corroborate confessions with independent other evidence (Escobedo v. Illinois, 1964), legal precedent continues to condone tactics that can induce innocent suspects to confess (e.g., Oregon v. Mathiason, 1977).

Of particular importance was the U.S. Supreme Court’s ruling in Frazier v. Cupp (1969) in which it suggested that trickery and deceit, as in misrepresenting the evidence, does not put an innocent person at risk to confess (in this case, police told Frazier that an associate had implicated him, which was not true). This ruling is interpreted as authorizing police to extract confessions by lying to suspects about the evidence. To this day, despite numerous documented false confessions taken in this way, and despite psychological research on misinformation effects, the Court has not revisited this issue. Consequently, confessions extracted through police deception are rarely suppressed despite the confluence of evidence from psychological science and catalogued exonerations resulting from false confessions.

Minimization tactics aim to provide the suspect with moral justification and face-saving excuses for the crime in question (e.g., that the actions were provoked or otherwise justified). In the 1897 case of Bram v. United States, the Supreme Court weighed in on this tactic and held that a confession “must not be extracted by any sort of threats or violence, nor obtained by any direct or implied promises, however slight” (pp. 542–543). Since that time, however, U.S. courts have consistently disregarded Bram and set it aside (see Arizona v. Fulminante, 1991).

In summary, innocents are confronted with legally permissible yet manipulative interrogation approaches such as minimization-theme and false-evidence ploys during custodial interrogations. Both basic and applied psychological science demonstrates, and naturalistic evidence confirms, that these techniques reliably increase innocents’ willingness to falsely confess, especially those from dispositionally vulnerable groups. At this stage, the cumulative disadvantages produce a false confession that, as we describe next, has the potential to undermine the ensuing investigation.

Confession effects on verdict decisions

When a suspect who confesses to police recants that confession, often immediately they will attempt to argue that it was coerced despite Miranda, sometimes plead not guilty, and go to trial. Sometimes a pretrial suppression hearing is held in which the judge determines whether the confession was voluntary by law and admissible as evidence. Confessions are overwhelmingly deemed voluntary at this stage primarily because the suspects waived their Miranda rights (Kassin, Scherr, & Alceste, 2019). Hence, a consequence of the triggering stage of the cumulative-disadvantage framework—a Miranda waiver—facilitates the credibility of evidence much later by perpetuating subsequent disadvantages such as false confessions. Confessions ruled voluntary are admitted at trial and presented to the jury, sometimes accompanied by a special instruction. But can juries overcome their commonsense beliefs that innocent people do not confess to crimes they did not commit?

Over the years, mock-jury studies have shown that confessions have a substantial impact on judgments of guilt—a greater impact, for example, than eyewitness and character testimony (Kassin & Neumann, 1997). People do not adequately discount confession evidence even when the confessions are perceived to have been coerced by police (Kassin & Sukel, 1997); even when told that the defendant suffers from a mental illness or was under duress (Henkel, 2008); even when the defendant is a juvenile (Redlich, Ghetti, & Quas, 2008; Redlich, Quas, & Ghetti, 2008); even when the confession was reported secondhand by an informant motivated to lie (Neuschatz, Lawson, Swanner, Meissner, & Neuschatz, 2008; Neuschatz et al., 2012); and, even at times, when the confession is flat-out contradicted by exculpatory DNA (Appleby & Kassin, 2016). The disadvantages are so compelling at this stage that people perceive coercive interrogation tactics that elicit a confession as more acceptable, and the confession as more voluntary, when other incriminating evidence (that may or may not be independent of the confession; see above) suggests the defendant’s guilt (Greenspan & Scurich, 2016; Shaked-Schroer, Costanzo, & Berger, 2015).

Perhaps it is not surprising that confession evidence biases inexperienced lay triers of fact. But judges are similarly affected. In one study, experienced judges read a case summary with strong or weak evidence and a confession elicited by high- or low-pressure interrogation tactics. Rationally, judges were less likely to see the confession as voluntary when it resulted from a high-pressure interrogation compared with a low-pressure interrogation (29% vs. 84%, respectively). Yet even the high-pressure confession, deemed involuntary, significantly increased the percentage of guilty verdicts. In the weak-evidence condition, which yielded a mere 17% conviction rate without a confession, a significant increase in convictions was produced not only by the low-pressure confession (96%) but also by the high-pressure confession (69%). No differently than mock juries, judges were so captivated by the confession that they did not disregard it when it was in their own view coerced and they were legally required to do so (Wallace & Kassin, 2012). Hence, by the fourth stage of the cumulative-disadvantage framework, the disadvantages are so impressive as to compel highly educated and presumably neutral legal professionals to render guilty verdicts.

Confession effects on plea bargains

Although the U.S. Constitution guarantees defendants the right to be tried by a jury of their peers, and although it is the trial that serves as the theatrical centerpiece of the legal system, the vast majority of U.S. cases (and around the world) are resolved via plea bargains. Although estimates vary slightly depending on the type of crime, approximately 95% of cases in U.S. state, federal, and juvenile courts are resolved using pleas (Redlich, Bibas, et al., 2017). For innocent suspects brought to this stage, cumulative disadvantages of prior experiences weigh on the process.

As an empirical matter, suspects who waive their rights, get interrogated, and confess are more likely to have their cases resolved through plea bargains than go to trial compared with those who do not confess (Leo, 1996a; Redlich, Yan, Norris, & Bushway, 2018). A guilty plea is a conviction, one that was decided without the benefit of the adversarial process (Newman, 1966). Thus, adding to the list of growing disadvantages at this point, another manner in which false confessions, validated by a presumed voluntary Miranda waiver, have corruptive effects is that they lead defendants to plead guilty. When that happens, defendants forgo most due process rights otherwise afforded—including but not limited to the presumption of innocence, the right to have guilt proven beyond a reasonable doubt, the right to confront and cross-examine their accusers, and the right to a jury of their peers (Redlich, Bibas, et al., 2017).

Two often-waived rights in particular can have severe ramifications that further illustrate the sheer magnitude of the accumulated disadvantages for innocents who go on to plead guilty. First, by pleading guilty, many defendants are forced to give up the right to appeal their conviction. For innocents, this waiver virtually eliminates the possibility of future exoneration because they are considered to have voluntarily relinquished their opportunity to use postconviction appeals to rectify a wrongful conviction. In no uncertain terms, the fate of these innocents is sealed because of an accumulation of disadvantages that caused them to plead guilty. Second, defendants who plead guilty may knowingly or unknowingly relinquish the right to full discovery of the evidence the state has against them. In the context of trials, the state must disclose to the defense all exculpatory and material evidence; the failure to do so is considered prosecutorial misconduct. In the context of pleas, however, the Supreme Court in United States v. Ruiz (2002) decided that “exculpatory impeachment evidence” (i.e., evidence that speaks to the credibility of witnesses) does not have to be disclosed. A federal court recently extended this ruling to traditional exculpatory material (Alvarez v. The City of Brownsville, 2018).

For defendants who are innocent, the effect of waiving discovery mirrors the effect of waiving the right to appeal because innocent defendants, who are more likely to have exculpatory evidence than guilty defendants, are often not aware of the state’s evidence favoring their innocence until well after they have accepted the guilty plea. The fate of the innocent person who pleads guilty is virtually sealed, further handicapping the possibility of exoneration. Why, then, would innocents decide to plead guilty?

The harmless-error myth

In Arizona v. Fulminante (1991), the U.S. Supreme Court ruled that an erroneously admitted confession does not necessarily entitle a convicted defendant to a new trial. Invoking the principle of “harmless error,” the Court ruled that appeals courts reviewing confession-based convictions must determine, first, whether a trial error occurred and, second, whether that error was prejudicial or harmless. In other words, the Court stated that even if a confession was coerced and should have been excluded from trial, the conviction could stand if other evidence was so compelling that the jury would still have found the defendant guilty beyond a reasonable doubt. Thus, if an appeals court sees other corroborating evidence of guilt, then the erroneously admitted confession is deemed harmless (Kassin, 2012).

This harmless-error doctrine applied to confessions is fatally flawed because it rests on the mistaken core assumption that the alleged other evidence is independent of the confession that was coerced and should have been excluded. Analyses of postconviction appellate reviews demonstrates the problem. According to Garrett (2010), appellate courts that had reviewed cases involving several confessors who were later exonerated had affirmed their convictions by citing the “overwhelming nature of the evidence against them” (p. 1107). As noted earlier, empirical research shows that confessions can bias eyewitnesses, alibi witnesses, polygraph judgments, handwriting judgments, and other forms of forensic evidence (Kassin, 2012; Kassin et al., 2013). In short, harmless error is a myth suggesting that the courts are oblivious to forensic confirmation biases and the cumulative disadvantages that create an illusion of corroboration, oftentimes thwarting an innocent defendant’s attempts at postconviction relief.

Persisting stigma after exoneration

After exoneration, research has found that the damaging influence of a confession persists, thus continuing to accumulate disadvantages. Doug Warney was an intellectually disabled man with mental health issues when he was convicted of murder. His 1997 conviction was based on a richly detailed confession to police during 12 hr of interrogation. After 9 years in prison, Warney was exonerated by DNA and released. Yet when he sought reparations through the state’s compensation statute, a court ruled that he was ineligible to do so because his wrongful conviction was his fault, resulting from his “own conduct”—namely, the police-induced false confession. The American Psychological Association filed an amicus brief in this case as well (American Psychological Association, 2010). Fortunately for Warney, the New York State Court of Appeals unanimously decided in his favor, disagreeing that Warney contributed to his own wrongful conviction.

Adding insult to injury, the story of Doug Warney illustrates one of the many ways in which the cumulative disadvantages that accrue to people who were wrongfully convicted by confession often persist even after they are exonerated and released back into the community. In theory, the path from release (i.e., the point in which individuals leave prison) to exoneration should be certain and clear-cut. For many innocents, this idea holds true and they are officially exonerated on the day of release (Scherr & Normile, 2019). For those who have falsely confessed, however, the path can be tortuous. Among cases involving serious crimes (e.g., murder and sexual assault) catalogued in the National Registry of Exonerations, individuals whose wrongful conviction involved a false confession (compared with those that did not) experienced an average delay of over a year between their release and official exoneration (Scherr & Normile, 2019). Why? A primary reason is that prosecutors persevere in their beliefs about the false confessor’s guilt, often concocting unrealistic theories to reconcile DNA and other forms of contradictory evidence (Appleby & Kassin, 2016). In some instances, this refusal to confront unimpeachable evidence of innocence after confession has drawn negative commentary from the news media (e.g., Martin, 2011) and the label “innocence deniers” (Bazelon, 2018).

Mercifully, some wrongly convicted confessors who suffered the accumulated disadvantages are exonerated—consistent with a general trend of wrongful exonerations that are being discovered at alarming rates around the globe. Exoneration is a fortunate experience for those wrongly convicted, but it masks the difficulties that many experience when rematriculating back into society (Thompson, Molina, & Levett, 2012; Westervelt & Cook, 2012; Westervelt & Humphrey, 2001). Only recently has psychological science started to identify the stigma that some innocents face after exoneration (Clow & Leach, 2013, 2015). The enduring stigma that follows from conviction can be conceptualized as an extension of basic stigma-by-association processes whereby unfavorable judgments of a disfavored group, such as guilty criminals, are linked to target individuals who do not belong but are nonetheless associated with that stigmatized group (Molet, Stagner, Miller, Kosinski, & Zentall, 2013; Pryor, Reeder, & Monroe, 2012).

Exonerees who falsely confessed are particularly at risk of being stigmatized, which is consistent with the cumulative-disadvantage framework and further extension of the disadvantages postconviction. Primarily because false confessions are counterintuitive and hard to understand as a matter of common sense, people fail to fully appreciate how situational factors can lead innocents to confess (e.g., Chojnacki, Cicchini, & White, 2008; Henkel, Coffman, & Dailey, 2008; Kassin, 2017; Leo & Liu, 2009; Mindthoff et al., 2018; Woestehoff & Meissner, 2016). Over a wide range of contexts, research has shown that people in Western cultures routinely commit the fundamental attribution error (Ross, 1977, 2018), or correspondence bias (Gilbert & Malone, 1995), making personal attributions for other people’s actions while underestimating the role of situational factors. In the context of our framework, observers are prone to blame false confessors for their fate (“I would never falsely confess”) and question their culpability (“Why else would they confess if they were not guilty?”). Thus, false confessors become associated with—and continue to be stigmatized like—those who are rightfully convicted.

Compared with those wrongfully convicted on the basis of mistaken eyewitness identifications or other forms of evidence, those wrongly convicted by confession are more likely to be perceived as guilty and less likely to be seen as deserving of government assistance—even after exoneration (Clow & Leach, 2015). The enduring negative stigma associated with falsely confessing also undermines people’s willingness to financially compensate these innocents (Kukucka & Evelo, 2019; Luna & Kieckhaeifer, 2018). False confessions can be so influential as to precipitate a series of negative judgments of these individuals as lacking intelligence, suffering from mental health issues, not entirely innocent, and, ultimately, less deserving of government-sponsored reintegration aids such as psychological and career counseling and job training (Scherr, Normile, & Putney, 2018). Along with replicating the undermining series of negative judgments associated with wrongful convictions based on false confessions, the same series of judgments follows after innocents offer false guilty pleas (Scherr et al., 2019).

We highlight this nascent area of research to bring attention to the importance of the exoneration experience and the fact that many false confessors and false guilty pleaders continue to be stigmatized. This research dispels the commonly held belief that when the legal system corrects its errors, the victims of injustice become whole and seamlessly reintegrate into society. As the story of Doug Warney illustrates, this assumption is incorrect. Indeed, “Just as commonsense folk psychology may ‘blind’ people from seeing innocence after confession, it may also keep them from accepting innocence even after exoneration” (Kassin, 2017, p. 958). Showing the lasting effects that an accumulation of disadvantages presents, several U.S. jurisdictions have provisions that virtually eliminate the possibility for some exonerees to be eligible for reintegration services and compensation. Referred to as “contributory provisions,” these rules stipulate that in order for exonerees to be eligible for aid and compensation they could not have contributed to their own conviction. Contributory provision stipulations are typically interpreted in ways that prevent exonerees who confessed or pled guilty from access to financial support and services. In these instances, the burden is on them to prove that they did not contribute to their own wrongful conviction.

To hold innocent people who confess or plead guilty under pressure responsible for their own wrongful conviction is a form of victim blaming that indicates just how poorly understood these phenomena are throughout the legal system. Echoing prior calls for reforms (e.g., Scherr et al., 2018; Shlosberg, Mandery, West, & Callaghan, 2014), we believe that all officially exonerated individuals should be guaranteed access to aid and compensation—without exception. Partitioning out aid on a case-by-case basis allows the biases that contributed to the cumulative disadvantage in the first place to remain a part of the decision-making process and preclude this large class of exonerees from gaining access to critical reintegration services.

To summarize, innocents who have waived their rights, falsely confessed, and been wrongly convicted via jury verdict or false guilty pleas continue to have difficulties both before and after exoneration. First, their path to exoneration is harder; second, they continue to be stigmatized and perceived as not entirely innocent. The disadvantages are especially apparent among innocents who reside in jurisdictions that have contributory provision statutes preventing these exonerees from accessing guaranteed aids and services because they are perceived as responsible for their own wrongful conviction.

Removing the requirement to self-invoke

Two of the U.S. Supreme Court rulings discussed earlier—Berghuis v. Thompkins (2010) and Florida v. Powell (2010)—underlie our first policy recommendation that suspects should not be required to self-invoke their interrogation rights (Smalarz et al., 2016). Over the years, the courts have weakened Miranda requirements through varying loopholes and exceptions (Kassin et al., 2019; Weisselberg, 2017). To restore the safeguards in the Constitution as intended, interrogations must be preceded by an expectation that suspects have and want to exercise their rights. Presuming by default that suspects prefer silence and legal counsel—rather than the manipulative dynamic that currently exists—removes the unreasonable requirement that suspects must speak up, or self-invoke, to remain silent. In seeking reform, we roll the clock back 50-plus years to the Miranda Supreme Court, which described the atmosphere of police interrogations as “inherently compelling” and sought to protect people from it—per the Fifth and Sixth Amendments.

The need to activate Miranda without self-invocation is particularly needed for suspects who are innocent and who do not hesitate to waive their rights. In addition, the courts should ensure that warnings are administered, not circumvented, in a timely manner; that these warnings be stated in a language that is simple, accessible, and not manipulative; that adolescents and other vulnerable individuals be protected; and that the invocation of rights is by default accepted unless explicitly waived according to the knowing, intelligent, and voluntary criteria in Miranda.

Video recording the interrogation

Echoing calls from scientists, legal scholars, policymakers, and many practitioners, all jurisdictions should require, without exception, that all interrogations be video recorded in their entirety—including all precustodial interviews, the administration of Miranda rights, and custodial interrogations—in addition to the confession itself (see Kassin & Thompson, 2019).

At present, half of all states, the District of Columbia, and federal agencies require the full recording of custodial interrogations; other states do not. Note that interview data indicate that law-enforcement agencies that have moved to recording interrogations become highly supportive of the practice (Sullivan, 2004, 2008).

We believe that the empirical benefits of recording are substantial. In one study, experienced police officers interrogated suspects who were guilty or innocent of a mock theft. Some were informed that their session would be recorded; others were not. Results showed that camera-informed police participants were less likely than those uninformed to use high-pressure interrogation tactics and that suspects perceived them as trying less hard to get them to confess (Kassin, Kukucka, Lawson, & DeCarlo, 2014). In a second study, police officers investigated a mock crime scene, interrogated two innocent suspects, and filed an incident report about the interrogations. Results showed that police understated their use of various tactics in their reports relative to what the actual tapes of the interrogation sessions (which were secretly recorded) showed. Hence, observers who later read an interrogation report versus a verbatim transcript perceived the process to be less pressured; they were also somewhat more likely to misjudge suspects as guilty (Kassin, Kukucka, Lawson, & DeCarlo, 2017). In short, video recording has two beneficial effects: First, the mere presence of a camera, an accountability cue, will inhibit interrogators from using highly manipulative tactics and encourage instead the kinds of cognitively based interviewing practices that are starting to reshape modern-day police work; and second, for police, prosecutors, judges, and juries, the recordings preserve an accurate memorial account of the exchange between the interrogator and suspect, providing insight into the Miranda waiver, the voluntariness of the process, and the reliability of the statement that is produced.

To this day, law-enforcement opponents still argue that recording will distract or inhibit suspects from talking to police. This concern is unfounded. In a fully randomized field experiment, researchers analyzed 122 real cases in which the suspects were randomly informed or not informed that their sessions would be recorded. The results were clear: Camera-informed suspects spoke as often as uninformed ones; they were as likely to waive their rights; they were as likely to make admissions, not just denials; and detectives rated them as equally cooperative (Kassin et al., 2019).

Curbing trickery and deceit

Heavily influenced by the Reid technique, custodial interrogations in the United States are rife with tactics that are legally permissible yet psychologically coercive. Informed by psychological science, as described earlier, two tactics are notably problematic for all suspects, particularly those who are innocent: the presentation of false evidence and minimization themes that imply leniency. As the story of Marty Tankleff illustrates, there is no shortage of horrific examples to illustrate how this type of misinformation can lead innocent people to confess—and to plead guilty to crimes they did not commit (Wynbrandt, 2016).

Now that there is a robust body of psychological science and compelling anecdotal evidence, the time has come for U.S. courts to refuse to admit confessions elicited by outright lying to suspects about the evidence. The U.S. Supreme Court has not revisited this issue in 50 years (Frazier v. Cupp, 1969). This practice is prohibited in most Western countries; the consensus within the scientific community concerning the risk is clear. In a survey of 87 PhD confession experts from all over the world, 94% agreed that “presentations of false incriminating evidence during interrogation increase the risk that an innocent suspect would confess to a crime he or she did not commit,” and 100% agreed that “misinformation about an event can alter a person’s memory for that event” (Kassin, Redlich, Alceste, & Luke, 2018).

U.S. courts are out of step with psychological science when it comes to this form of trickery and deception. It should be banned. Police then adapt to this change by removing this particular weapon from their confrontation-based tool box or by shifting to a more cognitive, information-gathering approach known as investigative interviewing. Investigative-interviewing approaches are widely favored in Europe (see Bull, 2014; Vrij, Mann, Kristen, & Fisher, 2007; D. Walsh, Oxburgh, Redlich, & Myklebust, 2016; Williamson, 2006) and produce interrogation outcomes that are more diagnostic of guilt and innocence (Meissner et al., 2014) and are supported by the U.S. government’s high-value detainee interrogation group research program (Brandon, 2011; Meissner, Surmon-Böhr, Oleszkiewicz, & Alison, 2017).

Blind testing of forensic examiners

A robust body of evidence now shows that the forensic sciences are not infallible but rather are subject to bias and error (e.g., Dror, 2016; Dror & Murrie, 2018; Kukucka, 2018; Saks et al., 2003). The problem of biased-induced errors has become so well established that the President’s Council of Advisors on Science and Technology (2016) issued a report emphasizing the practice of blinding examiners to potentially biasing contextual information. One source of error stems from confirmation biases in which examiners rely on information that is consistent with their preconceived expectations (Friedrich, 1993; Kassin et al., 2013; Nickerson, 1998). Strong evidence, such as confessions, prompt biased-induced errors that can distort forensic-analysis interpretation (Kukucka, 2018).

Our cumulative-disadvantage framework highlights the importance of mandating blind testing among forensic examiners—including mental health practitioners who serve as forensic experts. Imagine an interrogator telling a forensic examiner that the suspect had confessed. That contextual information may set off a series of cognitive processes whereby the forensic examiner, exposed to biasing information, focuses on data points that are inculpatory and disregards those that are exculpatory. To make matters worse, surveys of forensic science and mental health practitioners provide strong evidence for a “bias blind spot” (Kukucka, Kassin, Zapf, & Dror, 2017; Zapf, Kukucka, Kassin, & Dror, 2018).

One solution to minimize the effect of confessions and other strong forms of contextual information is to require linear sequential unmasking protocols that regulate the amount of information examiners receive from one point in time to another by ensuring that they have access to task-relevant evidence (e.g., there is no reason for latent fingerprint experts to receive exposure to a confession when their task involves matching two or more visual stimuli; Dror et al., 2015). Taking such an approach allows examiners to gain access to contextual information, when essential, and update their judgments so long as the revisions are documented and transparent. The initial response from laboratories that have adopted such protocols is encouraging with the general sentiment that implementation was relatively seamless, and the examiners’ reports and judgments are viewed with more confidence (Archer & Wall-man, 2016; Found & Ganas, 2013). Also of note is a second-level suggestion (Wells, Wilford, & Smalarz, 2013)—now supported by empirical research—that forensic examiners whose task it is to make pattern-match judgments do so by exposure to filler items modeled after an eyewitness lineup (Quigley-McBride & Wells, 2018).

Guilty pleas—rules of disclosure

All legal decisions—including guilty pleas—must be made knowingly, intelligently, and voluntarily (Redlich, 2016). In our estimation, it is not possible to make a truly knowing and intelligent decision without full knowledge of the evidence. In United States v. Ruiz (2002), the U.S. Supreme Court argued that exculpatory impeachment evidence is relevant to the fairness of a trial but not the voluntariness of a guilty plea. Can an unknowing plea be truly voluntary? Citing numerous state and federal cases as precedent, the Fifth Circuit Court of Appeals recently held that withholding traditional exculpatory material in the plea context does not violate a defendant’s due process rights (Alvarez v. The City of Brownsville, 2018). Only 14 states have discovery rules that explicitly address pleas—and of these, only 10 specifically cite the discovery of exculpatory material (a few have case law on this issue; e.g., Buffey v. Ballard, 2015; see Zottoli et al., 2019). Withholding exculpatory evidence is a consistent contributor to wrongful convictions. In the context of confessions, it is likely that the defendant knows that he or she confessed. However, he or she may not know the possible weaknesses of that confession (e.g., the presence of factual errors; contamination). Without a video recording and the turning over of the video recording to the defense, these weaknesses may remain obscured and unchallenged.

One way to limit the accumulation of disadvantages is through open-file discovery rules that are commonly recommended to combat prosecutorial misconduct and subsequent wrongful convictions (Alkon, 2014, 2017). Jurisdictions with open-file discovery policies essentially turn over the entire case file to the defense. Seventeen states currently have such policies (Turner & Redlich, 2016). In comparing an open-file-policy state (North Carolina) to a closed-file-policy state (Virginia), researchers found that these discovery policies extended to the provision of preplea evidence as well. More specifically, prosecutors and defense attorneys in North Carolina reported turning over and receiving significantly more preplea evidence than their counterparts in Virginia (Turner & Redlich, 2016). Among mock prosecutors in experimental lab studies, those taking open-file approaches turned over more discovery material and more exculpatory evidence than those not taking open-file approaches. In contrast, mock prosecutors informed about the Ruiz decision turned over significantly less discovery material and fewer pieces of exculpatory evidence than those in the control condition (Luna & Redlich, 2020).