Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

False confessions are not a new or novel phenomenon. From colonial Salem through the twenty-first century; in countries all over the world; in criminal justice, military, and corporate settings; many innocent people have confessed to crimes they did and would not commit. Within psychology, Munsterberg (1908) wrote about “untrue confessions” more than a hundred years ago; in On the Witness Stand, Bem (1966) and Zimbardo (1967) provided the first social psychological perspectives in the 1960s. Kassin and Wrightsman (1985a, 1985b) introduced a taxonomy consisting of three types of false confessions that served as a conceptual platform for current research. In light of this background, coupled with the recent wave of DNA exoneration s indicating the prevalence of false confessions, this chapter overviews the history of research in this area and then summarizes recent work specifically aimed at four questions: Why are innocent people often targeted for interrogation? Why do innocent people confess as a result of that process? Why do juries invariably believe false confessions—resulting in wrongful convictions that are later difficult to overturn? Finally, what can be done to prevent future miscarriages of justice caused by false confessions?

Historical Overview

In 1908, Harvard psychology professor Hugo Munsterberg published his precocious, controversial, and prescient book, On the Witness Stand: Essays in Psychology and Crime. Although the chapters of his book were not numbered, the sixth chapter, which was entitled “Untrue Confession s ,” spanned pages 135–171. Munsterberg’s chapter is fascinating for several reasons. First, he opened with a “sad story” of a man who was sentenced to death for murder based solely on a confession. Munsterberg was convinced that the man was innocent and that his confession was false; yet within 1 week of the man’s conviction, “he was hanged for a crime of which he was no more guilty than you or I” (p. 140). Interestingly, the case took place in Chicago—dubbed more than a hundred years later “The false confession capital” (CBS 60 Minutes, December 9, 2012). Second, Munsterberg clearly understood the potency of confession evidence in the courtroom because “it would be inconceivable that any man who was innocent should claim the infamy of guilt” (p. 142). Third, Munsterberg went on to discuss the Salem witch trials and speculate in general terms about the psychological causes of false confessions, using such words as hope, fear, promises, threats, suggestion, cunning calculations, passive yielding, real conviction, shock, fatigue, emotional excitement, melancholia, auto-hypnosis, dissociation, insanity, and self-destructive despair. Finally, it appears that Munsterberg’s opinion found its way into the Chicago newspapers, where he was roundly criticized in headlines that screamed about “Harvard’s Contempt of Court,” “Science Gone Crazy,” and “long-distance impudence.”

Munsterberg’s early insights did not spur action within psychology, a yet-to-become applied science, or within the law—even in the wake of Brown v. Mississippi (1936), a US Supreme Court opinion that banned third-degree interrogation tactics and articulated a measure of distrust in confessions. This sparked the development of explicit psychological approaches to interrogation—notably featuring the Reid technique first developed in the 1940s by criminologist Fred Inbau and Chicago police officer John Reid (entitled Criminal Interrogations and Confessions , the first edition of their manual was published in 1962; for an historical overview, see Leo, 2008).

Fifty years later, there was still only an occasional foray into the subject among social scientists. Putting his new self-perception theory to the test, Bem (1966) published an empirical article in the Journal of Personality and Social Psychology entitled “Inducing belief in false confessions” in which he offered a self-perception analysis in the laboratory of how saying (induced confession) can lead to believing (feelings of guilt). Then in the wake of Miranda v. Arizona (1966)—in which the US Supreme Court critically described psychological interrogation as “inherently coercive” and required police to apprise suspects in custody of their rights to silence and to counsel—Zimbardo (1967) published a social-psychological analysis of police interrogations in the inaugural issue of Psychology Today. At about the same time, occasional law review articles were published that offered “psychological” analyses of confessions—such as sociologist Driver’s (1968) “Confessions and the Social Psychology of Coercion,” published in the Harvard Law Review; and law professor Foster’s (1969) “Confessions and the Station House Syndrome,” which likened police interrogation to a trance-like state of hypnosis.

This was the spotty state of the pre-literature in 1978 when I defended my dissertation at the University of Connecticut (“Causal Attribution : A Perceptual Approach”), got married 1 week later, and traveled to Lawrence, Kansas for a postdoctoral fellowship to work with Larry Wrightsman. Larry, then Chair of the Psychology Department at KU, was studying jury decision-making—a burgeoning area of applied research for psychologists interested in the law. The opportunity to work with Larry provided one of those carpe diem moments in one’s life. Having been trained in basic social psychology, and interested in attribution theory, the application to juries, collaborating with Larry, was a natural fit (it did not hurt that when I arrived in town, Larry hosted a reception in his home, where he arranged for me to meet my intellectual hero, Fritz Heider—a professor emeritus whose classic 1958 book, The Psychology of Interpersonal Relations, inspired attribution theory; meeting Heider provided me a sense of closure, to use an old term from Gestalt psychology).

To get started, I proceeded to gather, solicit, pilot-test, and edit as many trial transcripts and videotapes I could get my hands on to develop a library of stimulus materials. An obvious pattern in these early returns could not be missed. It seemed that in trials that contained confessions in evidence, regardless of the circumstances, just about everyone voted guilty. Our first impulse was to edit out all confessions which, after all, reduced the variability of responses among mock jurors. Then we realized that this nuisance variable constituted a potent signal in jury decision-making—one that was often fraught with attributional ambiguity concerning a profound causal question: Why did the defendant confess—was he or she guilty or was the confession elicited by pressure from police?

In our first systematic jury studies, we discovered what we called a “positive coercion bias,” indicating that juries were more willing to discount the guilt implications of a confession when it was induced by threats of harm and punishment than by promises of leniency or other reward (Kassin & Wrightsman, 1980). In our second set of studies, we replicated this phenomenon and added that judge’s instructions as to what constitutes legal voluntariness and coercion did not keep jurors from convicting defendants induced to confess by promises (Kassin & Wrightsman, 1981). We later summarized this research in The American Jury on Trial: Psychological Perspectives (Kassin & Wrightsman, 1988).

Picking up on the history of scholarly interest in confessions within psychology, new developments were noteworthy on four fronts: (1) Miranda rights to silence and to counsel; (2) the social psychology of police interrogations; (3) the reformist movement in Great Britain; and (4) the Innocence Project ’s DNA exoneration cases.

Miranda: The Right to Remain Silent

In Miranda v. Arizona (1966), the US Supreme Court sought to protect suspects from conditions that might produce involuntary and unreliable confessions. Essentially, the court required police to inform suspects of their rights to silence and counsel. Only if suspects waive these rights “voluntarily, knowingly, and intelligently” can the statements produced be admitted into evidence. Miranda issues are often a source of dispute in the courts, particularly on the question of whether the warning-and-waiver requirement is sufficiently protective of the accused.

First and foremost is the concern that some number of suspects—because of their youth, lack of intelligence, lack of education, or mental health status—lack the capacity to understand and apply the rights they are given. Early on, Thomas Grisso reasoned that a person’s capacity to make an informed waiver requires various cognitive abilities. As described in his book, Juveniles’ Waiver of Rights: Legal and Psychological Competence, Grisso (1981) developed four instruments for measuring Miranda-related comprehension. Over the years, research with these instruments showed that adolescent suspects under age 15 do not comprehend their rights as fully or know how to apply them as well as older adolescents and adults (e.g., Grisso, 1998; Oberlander & Goldstein, 2001). This research soon morphed into a broader set of concerns about the interrogation of juveniles and the developmental risk of false confessions (Grisso et al., 2003; Owen-Kostelnik, Reppucci, & Meyer, 2006; Redlich & Goodman, 2003).

Following upon Grisso’s groundbreaking work, recent studies have focused on other aspects of Miranda and whether it sufficiently protects the accused. One line of research has shown that all warnings are not created equal. Rogers and his colleagues (2007) identified 560 different Miranda warning forms used by police throughout the U.S. and found that these warnings varied substantially in content, wording, and format. With regard to the effects of Miranda, and based on naturalistic observations of live and videotaped police interrogations, Leo (1996a) reported that four out of five suspects routinely waive the rights they are given and submit to questioning, in large part because police use various tactics designed to elicit the waiver. In an article entitled “Miranda’s Revenge,” Leo described this process as a confidence game (also see Leo & Thomas, 1998). Indeed, the waiver rate in the U.S. is similar to that found in Great Britain (e.g., Moston, Stephenson, & Williamson, 1992). In a third recent development, Rebecca Norwick and I proposed that innocence is a naïve mental state that leads people to trust the system, beginning with the decision to waive their Miranda rights. In a mock crime study, we found that participants who were truly innocent were far more likely than those who were guilty to waive their rights and agree to talk—81 to 36 %. The reason: They had done nothing wrong, had nothing to hide, and nothing to fear (Kassin & Norwick, 2004; Moore & Gagnier, 2008; for an overview, see Kassin, 2005). For an excellent review of the law and psychology surrounding Miranda, I refer you to one of Larry’s more recent books, The Miranda Ruling: It’s Past, Present, and Future (Wrightsman & Pitman, 2010).

Social Psychology of Police Interrogations

A second set of developments was initiated in 1985, when Larry and I edited a book on The Psychology of Evidence and Trial Procedure. In that edited volume, we wrote a chapter on “Confession Evidence” in which we overviewed the law, described common social influence practices of police interrogation, reviewed the scant research literature, and introduced a taxonomy that is still currently and universally used to distinguish three types of false confessions—voluntary, coerced-compliant, and coerced-internalized (Kassin & Wrightsman, 1985a, 1985b).

This three-part taxonomy of false confessions provided a starting point for understanding the psychological complexity of this counterintuitive phenomenon and the different types of influences that can put innocent people at risk to confess. By drawing on the literature on social influence, we distinguished, first, between the types of false confessions that arise when innocent people volunteer confessions without pressure (often to high-profile crimes that are in the news, the classic instance being the 200 false confessions volunteered to the 1932 kidnapping of Charles Lindbergh’s infant son) and those that come about through a process of police interrogation. Within the latter category, we then distinguished between cases in which innocent people, despite knowing they are innocent, transition from denial to confession as a mere act of compliance, to escape a harsh interrogation or because they are led to perceive that confession serves their own self-interest (when it comes to stress, discomfort, and the deprivation of need states, everyone has a breaking point) and those rarer instances of internalization in which innocent people, after having been subjected to highly misleading claims about the evidence, question their own innocence, come to infer that they were involved, and in some cases confabulate memories to support that inference (the baggage-heavy term “brainwashing” can loosely be used to describe this process).

Focused on American police interrogation tactics, as embodied in the popular Reid technique (Inbau & Reid, 1962; Inbau, Reid, Buckley, & Jayne, 2013), Kassin and McNall (1991) then identified two predominant clusters of tactics, referred to them by the terms maximization and minimization, and found that their presence in an interrogation leads people to infer varying degrees of consequence upon confession. Over the years, the US courts had ruled that confessions extracted by promises of leniency and threats of harm or punishment were not voluntary and, hence, not admissible in court. But what about the use of subtler, lawful tactics that covertly produce the same net effects on suspects’ expectations?

The specific problem that Kassin and McNall addressed was critical to the Reid technique in which interrogators minimize the seriousness of the crime through “theme development,” which provides the suspect with moral justification and face-saving excuses. Interrogators are thus trained to suggest to suspects that their actions were spontaneous, accidental, provoked, peer-pressured, drug-induced, or otherwise justifiable by external factors. In one of their studies, Kassin and McNall (1991) had subjects read transcript of an interrogation of a murder suspect. Three versions were produced in which the detective: (1) made a conditional promise of leniency, (2) used minimization by blaming the victim, or (3) used neither technique. Subjects read one version and estimated the sentence that they thought would be imposed on the suspect upon confession. The result: Minimization tactics led people to infer by pragmatic implication that leniency in sentencing will follow from confession—as if an explicit promise had been made. This research was conducted as part of Karlyn McNall’s undergraduate senior thesis. All of this early research was reviewed in a scholarly book, largely written by Larry, entitled Confessions in the Courtroom (Wrightsman & Kassin, 1993).

I should mention that Karlyn, a student at Williams College, became interested in the topic of false confessions because in 1984 a friend of her family, a young Berkeley student by the name of Bradley Page, was induced to confess to the murder of his girlfriend, Bibi Lee. This “confession”—a statement of speculation that Page gave when asked to imagine how he would have committed the crime—came after 16 h of relentless, guilt-presumptive, deception-filled interrogation in which Page was made to feel responsible for the death of his girlfriend; in which he was threatened with what would happen if he refused to confess; and in which he was misled into believing that he failed a polygraph, that his fingerprints were found at the crime scene, that his car was seen there by a witness, and that the killing appeared accidental. Page recanted his confession almost immediately, saying he was scared and confused. It was too late; the damage was done. In one trial the jury acquitted Page of first- and second-degree murder, but could not reach a verdict on the charge of voluntary manslaughter. In a second trial, a jury convicted Page of voluntary manslaughter. Ultimately, Page spent time in prison (for a description of this case, see Leo & Ofshe, 1998, 2001). It is worth noting, by the way, that the Bradley Page case is significant for another reason: On March 30, 1988, social psychologist Elliott Aronson testified as an expert for the defense on how someone could be induced to confess to a crime he did not commit (Davis, 2010; Tavris & Aronson, 2007).

Looking to study the “Milgramesque” interrogation tactics sanctioned by the Reid technique, my colleagues and I sought to develop an ethical laboratory paradigm that would both meet with IRB approval and confront innocent participants with a personally meaningful decision to confess. It was clear that entrapping people to cheat, steal, or otherwise commit an act that would embarrass them and cast them in a negative light would not be permitted. As part of her undergraduate honor’s thesis at Williams College, Katherine Lee Kiechel and I came up with an experimental paradigm that worked—the first to be used in the study of false confessions. What we came up with is now variously referred to as the computer crash or ALT key experiment in which the experimenter accused participants typing on a Dell desktop computer of causing the hard drive to crash by inadvertently pressing a key he had explicitly instructed to avoid. The experimenter, in his role, was upset—perhaps too much so (this experimenter went on to medical school). The first time we pilot-tested the paradigm, our participant, a female undergraduate, started to cry. We then dialed down the intensity of the experimenter’s reaction to the “crime,” however, and proceeded to run a study published in Psychological Science in which we found that people can be induced not only to sign a confession, but to internalize the belief in their own culpability and confabulate false memories when confronted with false evidence—in this study, a confederate witness claimed to have seen the participant hit the ALT key (Kassin & Kiechel, 1996). Soon replicated in other labs, as we will see shortly, this study served as a basis for a critical analysis of police-induced false confessions—and, in particular, the coercive effect of lying about evidence, which American police are permitted to do—which appeared in an American Psychologist article entitled “The psychology of confession evidence” (Kassin, 1997).

Developments in Great Britain

While my students and I were seeking to understand the social-psychological influences of police interrogations, a parallel set of important developments originated, across the Atlantic, led by Gisli Gudjonsson and his colleagues. A former police officer from Iceland, Gudjonsson is a clinical psychologist and professor of forensic psychology at the Institute of Psychiatry of King’s College in London. He had served as an expert in a number of high-profile false confession cases in England during the 1980s—including those of the “Guildford Four” and the “Birmingham Six.” Gudjonsson introduced the term “memory distrust syndrome” to help explain coerced-internalized false confessions (Gudjonsson & MacKeith, 1982) and he devised the popular Gudjonsson Suggestibility Scale , or GSS, to measure individual differences in susceptibility to influence during an interrogation (Gudjonsson, 1997).

Gudjonsson was at the forefront of a reform movement in the United Kingdom during the 1980s and 1990s that altered the nature of interrogation to make it less confrontational and required that these sessions be recorded. He has published a voluminous body of work, largely focused on individual differences in personality, mental health, and cognitive functioning and the tendency to confess or resist confession. The early work was comprehensively summarized in Gudjonsson’s (1992) The Psychology of Interrogations , Confessions , and Testimony, a highly influential book that was later supplanted by his 706-page The Psychology of Interrogations and Confessions: A Handbook (Gudjonsson, 2003). In light of Gisli’s British and clinical-personality perspectives, which contrast with my own American and social-cognitive perspectives, we have collaborated on several projects in an effort to provide a comprehensive overview of the literature as a whole (e.g., see Kassin & Gudjonsson, 2004, 2005; Kassin et al., 2010).

I should note that Gudjonsson’s early work coincided in time with the regulation of interrogations and confession evidence in England and Wales by the Police and Criminal Evidence Act of 1984 (PACE; Home Office, 1985). The most important interview procedures set out in PACE were that suspects who are detained must be informed of their legal rights; in any 24-h period, the detainee must be allowed a continuous period of rest of least 8 h; detainees who are vulnerable in terms of age or mental functioning should have access to a responsible adult; and all interviews shall be electronically recorded.

A few years later, in sharp contrast to the confrontational approach to interrogation characteristic of the American Reid technique, investigative interview practices in England became less confrontational. This new approach was developed through a collaboration of police officers, lawyers, and psychologists—such as Tom Williamson, Ray Bull, and Becky Milne. The acronym PEACE was used to describe the five distinct parts of the new interview approach (“Preparation and Planning,” “Engage and Explain,” “Account,” “Closure,” and “Evaluate”). PEACE continues to exist, with success, to this day—and has been adopted as an alternative to confrontation in Norway and New Zealand as well (for a review, see Williamson, 2006).

The Innocence Project’s DNA Exoneration Cases

Starting in the 1990s, historic developments outside of psychology conspired to cast a consciousness-raising spotlight on false confessions. In 1989, Gary Dotson became the first wrongfully convicted person to be exonerated by DNA testing. At a rapid pace, others soon followed. Three years later, the Innocence Project was founded by lawyers Barry Scheck and Peter Neufeld at the Benjamin Cardozo School of Law in New York to assist prisoners who could be proven innocent through DNA testing. Since that time, more than 300 people in the United States have been exonerated by DNA, including several who served time on death row. To everyone’s astonishment, false confessions have been a contributing factor in over 25 % of these wrongful convictions (www.innocenceproject.org/; Garrett, 2011).

Neither wrongful convictions in the U.S. nor the existence of false confessions as a problem was new (most notably, see Borchard, 1932). But in addition to the DNA exoneration s , other high-profile cases began to surface and were reported in books, newspapers, TV documentaries, and analyses of actual case files. Today, such names as Michael Crowe, Marty Tankleff, the West Memphis Three, the Norfolk Four, Christopher Ochoa, Juan Rivera, Eddie Joe Lloyd, Jeffrey Deskovic, and Gary Gauger dot the landscape of wrongful convicted confessors like celebrities. The horrific injustices they suffered now animate reform efforts.

This individual and aggregated case study approach can be seen in the writings, mostly published in law reviews, of sociologist Richard Ofshe; criminologist Richard Leo, his Ph.D. student; and Northwestern University law Professor Steven Drizin. In one article, Leo and Ofshe (1998) described 60 cases of proven, highly probable, and probable false confessions, which triggered a critique by Cassell (1999) in which he challenged the actual innocence of many of the confessors included in their sample, followed by a rejoinder by Leo and Ofshe (2001) entitled “The truth about false confessions and advocacy scholarship.” As measured by subsequent DNA and other exonerations, it is now clear that Cassell had denied the innocence of many innocents and that Leo and Ofshe were correct in most and possibly all of the “probable” and “highly probable” cases they had described.

Later focusing on a larger, more rigorous sample of proven exonerations, Drizin and Leo (2004) were able to describe the characteristics of 125 cases of proven false confession in the United States from 1971 to 2002. They found that 93 % of the false confessors were men. Overall, the vast majority of the confessions occurred in murder cases (81 %), distantly followed by rape (8 %) and arson (3 %). The most frequent bases of exoneration were that the real perpetrator was identified (74 %) or that new scientific evidence was discovered (46 %). Most recently, University of Virginia Law Professor Garrett (2010) compared 38 proven false confessions from the Innocence Project case files to the actual crime facts and found that 36 (95 %) contained accurate details about the crime that were not in the public domain—details that often served as a centerpiece for the prosecution at trial. This finding highlights the fact that most false confessions are contaminated with true facts about the crime that were known only to the perpetrator—and the police.

Taken together, the burgeoning literature in the United States and in Great Britain, in the laboratory and in actual cases, has shown that false confessions occur with some unknown frequency; that they share certain common features; and that many of the stories that accompanied these cases were incredible. Some of the cases that now breathe life into the study of false confessions are shocking and downright historic in their dimensions. The Central Park Five, a 2012 documentary, tells the tale of one such case. In 1989, a female jogger was raped, beaten, and left for dead in New York City’s Central Park. She managed to survive but could not remember anything about the attack—then or now. Within 72 h, five African- and Hispanic-American boys, 14–16 years-old, confessed to the brutal assault. Solely on the basis of these oral confessions, four of which were videotaped for all to see, the boys were convicted and sentenced to prison. Almost nobody questioned their guilt. The tapes themselves were compelling in that every one of the defendants described in vivid—though often erroneous—detail where and how the jogger was attacked and by whom.

Thirteen years later, Matias Reyes, in prison for three rapes and a murder committed subsequent to the jogger attack, stepped forward to admit that he was the Central Park jogger rapist and that he acted alone. I recall busily preparing for the upcoming semester in August of 2002 when I received a phone call from a producer at ABC Prime Time in which she asked if I could review the 1989 confessions in that case because ABC was about to get an exclusive interview with Reyes. I was stunned at the prospect that the Central Park jogger case was in doubt. Yet once I pored through the case files—the original police reports, suppression hearing and trial testimonies, confessions, and so on—and observed the new interview with Reyes; there was little doubt in my mind that the original confessions were false. After the ABC show aired on September 26, it was not clear if the city’s newspapers were willing to accept the new narrative on this heinous case and consider that the original defendants were innocent and wrongfully convicted. Nor was it clear how the Manhattan District Attorney’s office would react. Having reviewed the entire case, and uniquely positioned to educate the public on false confessions, I wrote an op-ed article entitled “False confessions and the jogger case,” which appeared in the New York Times on November 1 (Kassin, 2002).

Reinvestigating the case, the Manhattan district attorney questioned Reyes and discovered that he had accurate and independently corroborated guilty knowledge of the crime and that the DNA samples originally recovered from the victim’s body belonged to him. In a 58-page report released on December 5, the DA issued a report that dismantled the confessions and other evidence, noting that “Perhaps the most persuasive fact about the defendants’ confessions is that they exist at all” (p. 44). Two weeks later, on December 19, State Supreme Court Justice Charles Tejada vacated the original convictions. The Central Park jogger case now stands as a shocking demonstration of five false confessions resulting from a single investigation—an investigation conducted, I should add, not in a back alley of some small town when nobody was watching, but right in the heart of New York City at a time when the whole world was watching (for an overview, see Burns, 2011).

The Current Study of False Confessions

Picking up where Munsterberg left off, contemporary research on false confessions has analyzed various aspects of the confession-taking process and has relied on a range of methodologies. As noted earlier, one approach has involved a focus on actual case studies and aggregations of individual cases based on archived records. Other empirical methods include naturalistic observations of live and videotaped police interrogations; self-report surveys and interviews that purport to describe normative practices and beliefs; correlational studies that link various personal suspect characteristics and the tendency to confess; and controlled experiments—in laboratory and field settings—designed to assess police judgments of truth and deception, the effects of certain interrogation tactics on confessions, and the impact that confessions have not only on judges and juries but, more recently, on lay witnesses and forensic examiners. This literature is now sufficiently mature that it has served as the basis of a recent White Paper of the American Psychology-Law Society, only the second in its history (Kassin et al., 2010)—and a number of amicus curiae briefs submitted by the American Psychological Association (see Kassin, 2012).

The remainder of this chapter will provide a selective overview of the current empirical research literature on false confessions. In particular, I would like to frame this overview around four questions that must be asked about each and every case: (1) Why are innocent people interrogated in the first place? (2) What personal and situational forces lead innocent people confess to crimes they did not commit? (3) Why do judges, juries, and just about everyone else for that matter so often believe these statements? (4) What can be done to prevent future miscarriages of justice based on false confessions?

Why Are Innocent People Interrogated?

During the course of an investigation, police identify one or more suspects for interrogation. Sometimes, this identification is based on witnesses or other evidence, but often it is based on a judgment made during a pre-interrogation interview. In Criminal Interrogations and Confessions , Inbau et al. (2013) have for many years advised that the accusatory process of interrogation be preceded by an information-gathering interview designed to determine whether or not the suspect is lying, guilty, and in need of interrogation.

To determine if a suspect is telling the truth or lying, investigators are advised to ask a series of special “behavior provoking questions” and then observe changes in the suspect’s verbal and nonverbal behavior, noting pauses, changes in eye contact, facial expressions, postural shifts, fidgety movements, and other cues presumed to be diagnostic of deception. There is a dearth of solid research indicating that this technique produces high rates of accuracy at distinguishing truth and deception. In fact, research has consistently shown that most of the demeanor cues touted by the Reid technique do not empirically discriminate between truth-telling and deception (DePaulo et al., 2003). It is not surprising, therefore, that laypeople on average are only 54 % accurate and that police and other professionals perform only slightly better, if at all (Bond & DePaulo, 2006; Vrij, 2008).

In studies specifically aimed at evaluating the Reid approach to lie detection, the results are not impressive. Vrij, Mann, and Fisher (2006) had some participants not but others commit a mock crime they were motivated to deny. All participants were then interviewed using the Reid protocol. Results showed that substantive responses to the behavior-provoking questions did not significantly distinguish between truth tellers and liars in the predicted manner (e.g., the liars were not more anxious or less helpful). There is also no evidence to support the diagnostic value of the verbal and nonverbal cues that investigators are trained to observe. Kassin and Fong (1999) randomly trained some college students, but not others in the use of “behavioral symptoms” analysis cited by the Reid technique. All students then watched videotaped interviews of mock suspects, some of whom committed one of four mock crimes; others did not. Upon questioning, all suspects denied their involvement. As in so many experiments, observers could not reliably differentiate between the two groups of suspects. In fact, those who underwent training were significantly less accurate, more confident, and more biased toward seeing deception. Using these same taped interviews, Meissner and Kassin (2002) found that experienced police investigators exhibited the same low level of accuracy, albeit with higher levels of confidence.

As a natural consequence of the process whereby investigators make a judgment of truth and deception, research shows that police presume guilt when questioning suspects and that this guilt bias can lead them to engage in more aggressive interrogations with innocent suspects who vigorously deny involvement. To demonstrate, Kassin, Goldstein, and Savitsky (2003) had some participants but not others commit a mock crime, after which all were questioned by interrogators who by random assignment were led to presume guilt or innocence. Interrogators who presumed guilt asked more incriminating questions, conducted more coercive interrogations, and tried harder to get the suspect to confess. In turn, this more aggressive style made the suspects sound-defensive and led observers who later listened to the tapes to judge them as guilty, even when they were innocent. Follow-up research has replicated this effect, indicating the dangers of presuming guilt (Hill, Memon, & McGeorge, 2008; Narchet, Meissner, & Russano, 2011).

Why Do Innocent People Confess?

In recent years, false confessions have been uncovered in many cases involving innocent people who were wrongfully convicted—and these cases represent the tip of an iceberg. As a result of these miscarriages of justice, researchers have sought to identify two sets of risk factors: (1) dispositional risk factors indicating that some people are more malleable than others in an interrogation setting and more vulnerable to giving a false confession, and (2) situational risk factors indicating that some interrogation tactics in particular increase the likelihood that innocent people confess.

Personal Risk Factors

Some people are more vulnerable to influence than others—and at greater risk for false confessions. Focusing on personality traits, Gudjonsson (2003) has found that individuals who prone to compliance in social situations are especially vulnerable because of their eagerness to please others and a desire to avoid confrontation, particularly with those in authority. Individuals who are prone to suggestibility—whose memories can be altered by misleading questions and negative feedback—are also more likely to confess under interrogation. Most importantly, Gudjonsson notes that people who are highly anxious, fearful, depressed, delusional, or otherwise psychologically disordered are often at a heightened risk to confess under pressure.

Clearly, a suspect’s age and cognitive maturity is an important consideration. The 1989 Central Park jogger case described earlier illustrates the point, wherein five youths, 14–16 years old, were induced to give false confessions. In general, youths are overrepresented in the population of false confessions, thus suggesting that juveniles are at an increased risk in the interrogation room (see Drizin & Leo, 2004; Scott-Hayward, 2007). Criminal justice statistics are supported by a strong convergence of self-report studies and laboratory experiments (e.g., Grisso et al., 2003; Gudjonsson, Sigurdsson, & Sigfusdottir, 2009; Redlich & Goodman, 2003). These findings are consistent with basic research in developmental psychology indicating that adolescents are cognitively and psychosocially less mature than adults—exhibiting an “immaturity of judgment” that manifests itself in impulsive decision-making, a diminished focus on long-term consequences, and increased susceptibility to influence from external sources (for a review, see Owen-Kostelnik et al., 2006).

People with intellectual disabilities are also overrepresented (see Gudjonsson, 2003). Drizin and Leo (2004) identified at least 28 mentally retarded defendants in their sample of 125 false confessions and were quick to note that this 22 % likely underestimates the problem (intelligence test scores were not available or reported in most cases). Specifically addressing this problem, Appelbaum and Appelbaum (1994) note that people who are mentally retarded might confess to a crime merely to avoid the discomfort of police interrogation—that “Friendliness, as well as threats and coercion, can result in waivers and confessions” (p. 493).

With regard to tendencies toward compliance, people who are mentally retarded exhibit a high need for approval, particularly from others in positions of authority, which reveals itself in an acquiescence response bias (Shaw & Budd, 1982). A heightened suggestibility in response to misleading information, which can increase the risk of internalized false confessions, is also problematic. In studies conducted in England and the United States, respectively, Gudjonsson and Henry (2003) and Everington and Fulero (1999) found that people who are mentally retarded as a group score high on psychological measures of interrogative suggestibility, being more likely to yield to leading questions and to change their answers in response to mild negative feedback.

Situational Risk Factors

Just as some individuals are especially susceptible to giving false confessions, certain interrogation tactics can also increase the risk. If overused, three interrogation tactics in particular are problematic. The first concerns custody and interrogation time. Observational studies and police surveys in the United States have shown that most interrogations last from 30 min to 2 h (Feld, 2013; Kassin et al., 2007; Leo, 1996a, 1996b; Wald, Ayres, Hess, Schantz, & Whitebread, 1967). In contrast to these norms, police-induced false confessions are substantially longer. In their study of 125 false confessions, Drizin and Leo (2004) found, in cases in which interrogation time was recorded, that 34 % lasted 6–12 h, that 39 % lasted 12–24 h, and that the mean was 16.3 h.

A second tactic that can induce confessions from innocent people is the presentation of false evidence. In confronting suspects, American police will sometimes present false evidence of guilt (e.g., a fingerprint, eyewitness identification, or failed polygraph). In the United States, it is permissible for police to outright lie to suspects about the evidence. Yet empirical research clearly indicates that it can induce false confessions. There are two sources of evidence for this proposition. First, numerous proven false confessions featured the use of the false evidence ploy. In an illustrative case, 17-year-old Marty Tankleff was accused of murdering his parents, in 1989, despite the complete absence of evidence against him. Tankleff vehemently denied the charges for hours. Then his interrogator told him that his hospitalized father had emerged from his coma to say that Marty was his assailant (in fact, the father never regained consciousness and died shortly thereafter). Following this lie, and others, Tankleff became disoriented and confessed. Solely on the basis of that confession, he was convicted—a conviction that was not vacated until 2008 (Firstman & Salpeter, 2008).

The second source of evidence comes from basic psychology research showing that human malleability to influence through misinformation is broad and pervasive. By misrepresenting reality—through the use of confederates, bogus norms, false physiological feedback, counterfeit test results, and the like—one can substantially alter people’s visual perceptions, beliefs, behaviors, emotions, personal memories, and even certain medical outcomes, as seen in studies of the placebo effect. Studies specifically aimed at inducing false confessions have similarly shown that the presentation of false evidence increases the rate at which innocent research participants confess to prohibited acts they did not commit. In the first such study, Kassin and Kiechel (1996) accused college students typing on a keyboard of causing the computer to crash by pressing a key they were pre-instructed to avoid. Despite their innocence and initial denials, subjects were asked to sign a confession. In some sessions but not others, a confederate said she witnessed the subject hit the forbidden key. This false evidence nearly doubled the number of students who signed a written confession, from 48 to 94 %. Follow-up experiments have replicated the effect with different participant samples and under varying circumstances (Horselenberg et al., 2006; Horselenberg, Merckelbach, & Josephs, 2003; Perillo & Kassin, 2011; Redlich & Goodman, 2003).

A third problem concerns the use of minimization tactics. With suspects weakened by the highly confrontational stages of interrogation, interrogators are trained to minimize the crime through “theme development,” a process of providing moral justification or face-saving excuses, making confession seem like an expedient means of escape. Interrogators may suggest to suspects that their actions were spontaneous, accidental, provoked, peer-pressured, or otherwise justifiable by external factors. As described earlier, early studies showed that minimization tactics may lead people to infer that leniency in sentencing will follow from confession—even without an explicit promise (Kassin & McNall, 1991). In a laboratory paradigm more recently designed to elicit true and false confessions to cheating, Russano, Meissner, Narchet, and Kassin (2005) found that innocent participants were more likely to confess when leniency was promised than when it was not—and when minimization was used. In short, minimization provides police with a loophole in the rules of evidence by serving as the implicit but functional equivalent to a promise of leniency (which itself renders a confession inadmissible). The net result is to increase the rate of false confessions.

Does Innocence Put Innocents at Risk?

In 2005, I proposed a paradoxical hypothesis—that false confessions are facilitated not only by dispositional characteristics of weak and vulnerable suspects and situational aspects of custody and interrogation, but by the phenomenology of innocence. This hypothesis began with an observation, followed by a laboratory experiment indicating that innocent people are more likely than perpetrators to waive their Miranda rights to silence and to counsel (Kassin & Norwick, 2004; Moore & Gagnier, 2008). Additional research has since shown that innocent people behave in ways that are open and forthcoming in their interactions with police (Hartwig, Granhag, & Strömwall, 2007); offer up alibis freely, without regard for the fact that police would view minor inaccuracies with suspicion (Olson & Charman, 2012); and exhibit less physiological arousal on critical items of a concealed information polygraph test even when told about the crime (Elaad, 2011). Experiments in another context have shown that many participants who are accused of a transgression they did not commit—compared to those who are guilty—refuse to accept a plea offer, often to their own detriment, because they are confident of acquittal (Gregory, Mowen, & Linder, 1978; Tor, Gazal-Ayal, & Garcia, 2010).

Innocence as a mental state can have unpredictable effects on a suspect’s response to interrogation. Theorizing that innocence leads people to trust that justice will prevail, Jennifer Perillo and I examined the relatively benign bluff technique by which interrogators pretend to have evidence without further claiming that it implicates the suspect (e.g., stating that biological evidence was collected and sent for testing). The theory underlying the bluff is simple: Fearing the evidence to be processed, perpetrators will succumb to pressure and confess; not fearing that alleged evidence, innocents would not succumb and confess. Yet in two experiments, Perillo and Kassin (2011) found that innocent participants were substantially more likely to confess to pressing a forbidden key, causing a computer to crash, when told that their keystrokes had been recorded for later review. In a third experiment, innocent participants were more likely to confess to willful cheating when told that a surveillance camera had taped their session. Consistently, these participants noted that the bluff represented a promise of future exoneration, despite confession, which paradoxically made it easier to confess.

How Do Juries Perceive Confessions?

Confession evidence is devastating when presented in court. In fact, the power of confessions to influence juries is what led Larry and I to become interested in this type of evidence in the first place (Kassin & Wrightsman, 1980, 1981). When a suspect retracts a confession, pleads not guilty, and goes to trial, a judge determines at a pretrial suppression hearing whether the confession was voluntary and hence admissible as evidence. There are no simple criteria for making this judgment, but over the years the courts have ruled that whereas various forms of trickery and deception are permissible, confessions cannot be produced by physical violence, threats or harm or punishment, explicit promises of leniency, or interrogations conducted in violation of a suspect’s Miranda rights. Whatever the criteria, confessions ruled voluntary are admitted at trial. Hearing the admissible confession, the jury then determines whether the defendant is guilty beyond a reasonable doubt. But are people accurate and discriminating judges of confessions? The wrongful convictions of innocent confessors suggest a negative answer to this question.

To test whether police can distinguish between true and false confessions to actual crimes, Kassin, Meissner, and Norwick (2005) recruited male prison inmates from a corrections/facility in massachwsetts to take part in a pair of videotaped interviews. Each inmate gave both a true narrative confession to the crime for which he was incarcerated and a false confession to a crime he did not commit. Using this procedure, Kassin et al. compiled a videotape of ten confessions known to be true or false. College students and police investigators judged these statements, and the results showed that neither group exhibited significant accuracy, but that police were more confident in their judgments.

Over the years, mock jury studies have shown that confessions have a great impact on jury verdicts. This research has shown that confessions have more impact than eyewitness and character testimony (Kassin & Neumann, 1997) and that people do not adequately discount confession evidence even when the confessions are perceived to have been coerced by police (Kassin & Sukel, 1997; Kassin & Wrightsman, 1980); even when jurors are told that the defendant suffers from a psychological illness or interrogation-induced stress (Henkel, 2008); even when the defendant is a juvenile (Redlich, Ghetti, & Quas, 2008; Redlich, Quas, & Ghetti, 2008); even when the confession was given not by the defendant but by a second-hand informant who was motivated to lie (Neuschatz et al., 2012; Neuschatz, Lawson, Swanner, Meissner, & Neuschatz, 2008); and even when the confession is contradicted by exculpatory DNA that is explained away by the prosecutor (Appleby & Kassin, 2011).

In a study that well illustrates the potency of confession evidence, Kassin and Sukel (1997) presented subjects with one of three versions of a murder trial transcript. In a low-pressure version, the defendant was said to have confessed to police immediately upon questioning. In a high-pressure version, participants read that the suspect was in pain and interrogated aggressively by a detective who brandished his gun. A control version contained no confession in evidence. In some ways, participants presented with the high-pressure confession responded in a legally appropriate manner: They judged the statement involuntary and said it did not influence their decisions. Yet this confession significantly boosted the conviction rate. This pattern of results was recently replicated in a study of judges (Wallace & Kassin, 2012). In short, it appears that people are so influenced by confession evidence as a matter of common sense that they do not discount it when it is legally and logically appropriate to do so.

In actual cases, there is an additional reason why people place so much faith in confessions, even those that are false. Garrett (2010) recently compared 38 proven false confessions from the Innocence Project DNA exoneration case files to the actual crime facts and found that 36 (95 %) contained accurate details about the crime that were not in the public domain. Often the details served as a centerpiece for the prosecution at trial, with interrogating detectives testifying that these facts could have only been known by the perpetrator. The confessors in these cases were innocent, so they could not have possessed firsthand guilty knowledge. Thus, it appears that police had communicated these details—purposefully or not—through leading questions, factual assertions, photographs, or taking the suspect to the crime scene.

Illustrating that matters are even more complicated, Appleby, Hasel, and Kassin (2013) content-analyzed 20 false confessions and found that most contained vivid details about the crime and how, when, where, and why it was committed—including descriptions of the crime scene and the victim. Eighty percent also contained motive statements, often accompanied by a minimization theme that justified, excused, mitigated, or externalized blame. Half of the false confessors asserted that their statements were voluntary; 40 % expressed remorse; and 25 % apologized. In short, false confessions contain the kinds of cues that create an illusion of credibility.

What Consequences Follow from Confession?

Mock jury studies on the power of confession evidence are bolstered by archival analyses of actual cases, which show that approximately four out of every five proven false confessors who pled not guilty were convicted at trial (Drizin & Leo, 2004; Leo & Ofshe, 1998). These figures led Drizin and Leo to describe confessions as “inherently prejudicial and highly damaging to a defendant, even if it is the product of coercive interrogation, even if it is supported by no other evidence, and even if it is ultimately proven false beyond any reasonable doubt” (p. 959).

One reason confessions are so powerful is that it is hard to understand as a matter of common sense why anyone would confess to a crime he or she did not commit (Henkel, Coffman, & Dailey, 2008; Leo & Liu, 2009). When the statement also contains vivid details about the crime, the impact is even greater (Appleby et al., 2013). It now appears that the problems spawned by false confessions are even more troubling. In an article entitled “Why confessions trump innocence,” I reviewed research indicating that confessions, which elicit a strong inference of guilt, can bias witnesses and forensic examiners, thereby corrupting other evidence (Kassin, 2012). Hence, the wrongful convictions of innocent people who confess are based in part on the confessions themselves and in part on the cognitive confirmation biases triggered by these confessions (Kassin, Dror, & Kukucka, 2013).

In a demonstration of this point, Hasel and Kassin (2009) had participants witness a staged theft and then make an identification decision from a target-absent lineup. Two days later, they were given additional information and an opportunity to change their decision. When told that another suspect had confessed, 61 % of participants changed their initial decision and identified the suspect who had allegedly confessed. Those who were told that the individual they had identified confessed became more confident in their decision. Other researchers have shown that the belief in a suspect’s guilt can bias people’s judgments of inconclusive polygraph tests (Elaad, Ginton, & Ben-Shakhar, 1994), degraded speech recordings (Lange, Thomas, Dana, & Dawes, 2011), handwriting samples (Kukucka & Kassin, 2014), latent fingerprint samples (Dror & Charlton, 2006), and even complex DNA mixtures alibis (Marion et al., in press), (Dror & Hampikian, 2011).

A recent archival analysis is also consistent with the hypothesis that confessions can corrupt other evidence. Looking at the DNA exoneration files from the Innocence Project , Kassin, Bogart, and Kerner (2012) tested the hypothesis that confessions prompt additional evidentiary errors by examining whether other contributing factors were present in DNA exoneration cases containing a false confession. Sure enough, additional evidence errors were present in 78 % of these cases. Specifically, false confessions were accompanied by invalid or improper forensic science (63 %), mistaken eyewitness identifications (29 %), and snitches or informants (19 %). Consistent with the causal hypothesis that the false confessions had influenced the subsequent errors, the confession was obtained first rather than later in the investigation in approximately two thirds of these cases.

Proposed Reforms to Policy and Practice

When Larry and I started to explore the social psychology of confession evidence, there was no recognition of a problem to be solved and no reason for reform. In the wake of the DNA revolution, the discovery of numerous heart-wrenching false confessions, and psychological research explaining the mechanisms by which this can occur and how it can be prevented, there is now a wave of reform in the air. In that vein, many social scientists, legal scholars, and practitioners have recommended a policy that favors the electronic recording of entire suspect interviews and interrogations. In the recent American Psychology-Law Society White Paper noted earlier, Kassin et al. (2010) concluded with a strong recommendation for the mandatory electronic recording of interrogations. Currently, statutory provisions or supreme court rules require the recording of custodial interrogations for serious felonies in 20 states and the District of Columbia, with hundreds of other jurisdictions doing so on a voluntary basis (Sullivan, 2012). Importantly, the U.S. Department of Justice recently reversed its long standing opposition by establishing the presumptive requirement that the FBI and other federal law enforcement agencies record the custodial interrogations of felony suspects (Schmidt, 2014).

Still other recommendations for reform focus on the protection of vulnerable suspect populations (e.g., a requirement that minors be accompanied by a professional advocate, preferably an attorney) and a ban or limit on the use of certain police interrogation practices (e.g., the false evidence ploy and minimization tactics that imply leniency). In addition, the American Psychological Association has recently submitted a number of amicus curiae on the subject of confessions. Drawing upon the research these briefs’ stated that innocent people can be induced to confess through processes of interrogation, that judges and juries have difficulty assessing confession evidence, that the phenomenon of false confession is counterintuitive, and that psychological experts should be permitted to testify at trial because their testimony would draw from generally accepted research, and that it would assist the trier of fact (see Kassin, 2012).

Closing Thought

It’s now been over 100 years since Munsterberg (1908) had the audacity to dedicate a chapter of On the Witness Stand to “Untrue Confession s ”—a proposition no one at the time was prepared to accept. Seventy years later, not much had changed. Now, however, thanks to contributions from psychologists and other social scientists in the United States and abroad; reformers in Great Britain; the DNA revolutionaries of the Innocence Project ; countless individuals wrongfully convicted and imprisoned by false confessions; and courageous judges, lawyers, and lawmakers; there is now both the recognition that innocent people can be induced to confess to crimes they did not commit and a determination to prevent it from happening in the future. Larry Wrightsman has been an important part of this story.

Finally, I should mention that I do own and cherish a yellowed, frayed, hardbound copy of Munsterberg’s classic book. The title is embossed in gold block lettering. Several notes and comments are handwritten into the margins. Written on the front inside cover, in blue ink, there is also this personal note: “Dear Saul, This book is even older than our relationship!—Happy Birthday, Larry.” Back at ya, buddy—and thank you for everything.