R E G U L A R A R T I C L E

False Confessions: Causes,

Consequences, and Implications
Richard A. Leo, PhD, JD

In the past two decades, hundreds of convicted prisoners have been exonerated by DNA and non-DNA evidence,
revealing that police-induced false confessions are a leading cause of wrongful conviction of the innocent. In this
article, empirical research on the causes and correlates of false confessions is reviewed. After a description of the
three sequential processes that are responsible for the elicitation of false confessions—misclassification, coercion,
and contamination—the three psychologically distinct types of false confession (voluntary, compliant, and per-
suaded) are discussed along with the consequences of introducing false-confession evidence in the criminal justice
system. The article concludes with a brief discussion of the implications of empirical research for reducing the
number of false confessions and improving the accuracy of confession evidence that is introduced against a
defendant at trial.

J Am Acad Psychiatry Law 37:332– 43, 2009

In recent years, the media have reported numerous Nevertheless, the DNA exonerations provide a
high-profile cases in which individuals were con- window into the causes of erroneous prosecution and
victed of and incarcerated for serious crimes they did wrongful conviction. A disturbing number of these
not commit, only later to be exonerated.1 Many, cases involved false confessions given by innocent
though not most, of these exonerations occurred af- defendants during a psychologically coercive police
ter postconviction DNA evidence established inno- interrogation. In the Central Park Jogger case, for
cence of those convicted.2 In some of these cases, example, all five juveniles falsely confessed after
such as the Central Park Jogger case in New York lengthy unrecorded interrogations in which they
City, the DNA evidence established the innocence of were yelled at, lied to, threatened, and promised im-
multiple defendants who had been wrongly prose- munity in exchange for their admissions to partici-
cuted, convicted, and incarcerated. To date, more pating in the assault and rape.5 In 15 to 20 percent of
than 220 individuals have been exonerated by post- the DNA cases, police-induced false confessions were
conviction DNA testing and released from prison, the primary cause of the wrongful conviction.6,7
some from death row (e.g., see the Innocence Project Here too, however, the documented cases appear to
at www.innocenceproject.org). Although researchers represent the proverbial tip of the iceberg, as the
and scholars have long documented the problem of DNA exonerations again do not include most cases
wrongful conviction,3 the use of DNA testing to ex- in which there is no DNA to test. They also do not
include false confessions that were dismissed or dis-
onerate innocent prisoners and the sustained media
proved before trial, those that resulted in guilty pleas,
attention that it has received has increased public
those given for crimes that were not subject to post-
recognition that the criminal justice system often
conviction review (especially less serious crimes), and
convicts the wrong people.4 Although the number of those given in cases that contain confidentiality pro-
wrongful convictions continues to mount, the DNA visions (e.g., juvenile proceedings).
exonerations represent only a small part of a much False confessions raise important questions for so-
larger problem. For in most criminal cases, there was cial scientists, mental health professionals, policy-
and is no DNA evidence available for testing. makers, and the public. They are consistently one of
the leading, yet most misunderstood, causes of error
Dr. Leo is Associate Professor of Law, University of San Francisco, San in the American legal system and thus remain one of
Francisco, CA. Address correspondence to: Richard A. Leo, PhD, JD,
University of San Francisco School of Law, 2130 Fulton Street, San the most prejudicial sources of false evidence that
Francisco, CA 94117. E-mail: [email protected] lead to wrongful convictions. In this article, I will

332 The Journal of the American Academy of Psychiatry and the Law
 Leo

review and analyze the empirical research on the mented approximately 300 proven false confessions
causes and correlates of false confessions, the psycho- in recent decades.7 Researchers have also categorized
logical logic and various types of false confession, and cases involving likely, but nonproven, false confes-
the consequences of introducing false-confession ev- sions as highly probable or probable false
idence at trials. confessions.8
A false confession is an admission (“I did it”) plus Despite substantial documentation and analysis
a postadmission narrative (a detailed description of by scholars,8,9the phenomenon of police-induced
how and why the crime occurred) of a crime that the false confessions remains counterintuitive to most
confessor did not commit. Although other research- people. Most lay people believe in what has been
ers have also documented and analyzed numerous referred to as the myth of psychological interroga-
false confessions in recent years, we do not know how tion: that an innocent person will not falsely confess
frequently they occur. A scientifically meaningful in- to police unless he is physically tortured or mentally
cidence rate cannot be determined for several rea- ill.10 This belief has been noted by several scholars
sons. First, researchers cannot identify (and thus can- and documented in public surveys.7,11,12 The myth
not randomly sample) the universe of false of psychological interrogation persists because most
confessions, because no governmental or private or- people do not know what occurs during police inter-
ganization keeps track of this information. Second, rogations, and because they wrongly assume that in-
even if one could identify a set of possibly false con- dividuals do not act against their self-interest or en-
fessions, it is not usually possible as a practical matter gage in self-destructive behavior, such as falsely
to obtain the primary case materials (e.g., police re- confessing to a crime that they did not commit.
ports, pretrial and trial transcripts, and electronic re-
cordings of the interrogations) necessary to evaluate The Causes of False Confession:
the unreliability of these confessions. Finally, even in Misclassification, Coercion, and
disputed confession cases in which researchers are Contamination
able to obtain primary case materials, it may still be There is no single cause of false confession, and
difficult to determine unequivocally the ground there is no single logic or type of false confession.
truth (i.e., what really happened) with sufficient cer- Police-induced false confessions result from a multi-
tainty to prove the confession false. In most alleged step process and sequence of influence, persuasion,
false-confession cases, it is therefore impossible to and compliance and usually involve psychological
remove completely any possible doubts about the coercion.13 Police are more likely to elicit false con-
confessor’s innocence. fessions under certain conditions of interrogation,
Social science research on wrongful convictions, however, and individuals with certain personality
however, has demonstrated that there are four ways traits and dispositions are more easily pressured into
to prove a confession is false: (1) when it can be giving false confessions. To understand why criminal
objectively established that the suspect confessed to a suspects give false confessions, we must first under-
crime that did not happen (e.g., the presumed mur- stand how police investigators target criminal sus-
der victim is found alive); (2) when it can be objec- pects and how police interrogation works as a psy-
tively established that the defendant could not have chological process, before eliciting a suspect’s
committed the crime because it would have been admission and in the postadmission stage of
physically impossible to have done so (e.g., he was in interrogation.
another location at the time of the crime); (3) when There are three sequential errors, which occur dur-
the true perpetrator of a crime is identified and his ing a police-elicited false confession, that lead to a
guilt can be objectively established; or (4) when sci- wrongful conviction. Investigators first misclassify an
entific evidence, in recent years most commonly innocent person as guilty; they next subject him to a
DNA evidence, conclusively establishes the confes- guilt-presumptive, accusatory interrogation that in-
sor’s innocence.8 Despite these four possibilities, variably involves lies about evidence and often the
only a small number of alleged false confessions con- repeated use of implicit and explicit promises and
tain the independent case evidence or circumstances threats as well. Once they have elicited a false admis-
that allow the confessor to prove his innocence be- sion, they pressure the suspect to provide a postad-
yond dispute. Nevertheless, researchers have docu- mission narrative that they jointly shape, often sup-

Volume 37, Number 3, 2009 333

 False Confessions

plying the innocent suspect with the (public and at levels greater than chance; indeed, they routinely
nonpublic) facts of the crime. These have been re- make erroneous judgments.16,17 The method of be-
ferred to as the misclassification error, the coercion havior analysis taught by the police training firm
error, and the contamination error.7 Reid and Associates has been found empirically to
lower judgment accuracy, leading Kassin and Fong
The Misclassification Error to conclude that “the Reid technique may not be
The first mistake occurs when detectives errone- effective—and, indeed, may be counterproduc-
ously decide that an innocent person is guilty. As tive—as a method of distinguishing truth and decep-
Davis and Leo point out, “the path to false confession tion” (Ref. 17, p 512). The Reid-based Behavioral
begins, as it must, when police target an innocent Analysis Interview, which primarily consists of ask-
suspect. . . . Once specific suspects are targeted, po- ing 15 to 20 questions designed to evoke particular
lice interviews and interrogations are thereafter behavioral (verbal, nonverbal, and paralinguistic) re-
guided by the presumption of guilt” (Ref. 14, p 124). sponses from which the interrogator can allegedly
Whether to interrogate or not is therefore a critical discern whether a suspect is telling the truth or lying,
decision point in the investigative process. If police has its origins in the polygraph and like the poly-
did not erroneously interrogate innocent people, graph has been empirically shown to have high rates
they would never elicit false confessions. Because of error.7 As Kassin and Gudjonsson note,12 police de-
misclassifying innocent suspects is a necessary condi- tectives and other professional lie-catchers are accu-
tion for all false confessions and wrongful convic- rate only 45 to 60 percent of the time.
tions, it is both the first and the most consequential Innocent persons may be mistakenly targeted for
error that police make. suspicion and misclassified as guilty suspects for
There are many cognitive errors that lead police to other reasons. The suspect may, for example, simply
classify an innocent person mistakenly as a guilty be the most readily noticed person who fits a very
suspect. Perhaps the most prominent stems from general description given by an eyewitness or others.
poor and erroneous investigation training. American Although many may fit the description, the target
police are taught, falsely, that they can become hu- may be chosen simply because he happens to be no-
man lie detectors capable of distinguishing truth ticed by the police, reported by someone who had
from deception at high, if not near perfect, rates of seen a police sketch or falsely identified from a mug
accuracy. (The Chicago-based firm Reid and Associ- shot or lineup, or he fits an official profile of the
ates, for example, claims that detectives can learn to perpetrator. A suspect may also be targeted based on
discriminate truth and deception accurately 85 per- widespread crime-related schemas, including likely
cent of the time, although this rate seems to be rep- motives for the crime as well as the perpetrators likely
resented in their training seminars as 100 percent.7) to have such motives.18 Family members, for exam-
Detectives are misleadingly taught, for example, that ple, have been led to confess falsely to murdering
the subject who averts his gaze, slouches, shifts his wives, children, or parents, largely because police
body posture, touches his nose, adjusts or cleans his start with the assumption that most such murders are
glasses, chews his fingernails, or strokes the back of committed by family and proceed by ruling out fam-
his head is likely to be lying and thus is guilty.7 The ily before looking for other suspects.
subject who is guarded, uncooperative, and offers
broad denials and qualified responses is also believed The Coercion Error
to be deceptive and therefore guilty. However, social Once detectives misclassify an innocent person as
scientific studies have repeatedly demonstrated a guilty suspect, they often subject him to an accusa-
across a variety of contexts that people are poor hu- torial interrogation. Getting a confession becomes
man lie detectors and thus are highly prone to error particularly important when there is no other evi-
in their judgment about whether an individual is dence against the suspect, especially in high-profile
lying or telling the truth. Most people get it right at cases in which there is great pressure on police detec-
rates that are no better than chance (i.e., 50%) or the tives to solve the crime, there is no other source of
flip of a coin.15 Moreover, specific studies of police potential evidence to be discovered,19 and typically
interrogators have found that they cannot reliably there is no credible evidence against an innocent but
distinguish between truthful and false denials of guilt misclassified suspect. It is perhaps not surprising that

334 The Journal of the American Academy of Psychiatry and the Law
 Leo

most documented false confessions occur in homi- worse. The suspect may perceive that he has no
cides and high-profile cases.2,5 choice but to comply with the detectives’ wishes,
Once interrogation commences, the primary because he is fatigued, worn down, or simply sees no
cause of police-induced false confession is psycholog- other way to escape an intolerably stressful experi-
ically coercive police methods.20 Psychological coer- ence. Some suspects come to believe that the only
cion can be defined in two ways: police use of inter- way they will be able to leave is if they do what the
rogation techniques that are regarded as inherently detectives say. Others comply because they are led to
coercive in psychology and law, or police use of in- believe that it is the only way to avoid a feared out-
terrogation techniques that, cumulatively, cause a come (e.g., homosexual rape in prison). When a sus-
suspect to perceive that he has no choice but to com- pect perceives that he has no choice but to comply,
ply with the interrogators’ demands. Usually these his resultant compliance and confession are, by def-
amount to the same thing. Psychologically coercive inition, involuntary and the product of coercion.20
interrogation techniques include some examples of
the old third degree, such as deprivations (of food, Vulnerable Suspects
sleep, water, or access to bathroom facilities, for ex- Even though psychological coercion is the primary
ample), incommunicado interrogation, and induc- cause of police-induced false confessions, individuals
tion of extreme exhaustion and fatigue. In the mod- differ in their ability to withstand interrogation pres-
ern era, however, these techniques are rare in sure and thus in their susceptibility to making false
domestic police interrogations. Instead, when to- confessions.9 All other things being equal, those who
day’s police interrogators employ psychologically co- are highly suggestible or compliant are more likely to
ercive techniques, they usually consist of (implicit or confess falsely. Individuals who are highly suggestible
express) promises of leniency and threats of harsher tend to have poor memories, high levels of anxiety,
treatment. As Ofshe and Leo have written, “the mod- low self-esteem, and low assertiveness, personality
ern equivalent to the rubber hose is the indirect factors that also make them more vulnerable to the
threat communicated through pragmatic implica- pressures of interrogation and thus more likely to
tion” (Ref. 21, p 1115). Threats and promises can confess falsely.12 Interrogative suggestibility tends to
take a variety of forms, and they are usually repeated, be heightened by sleep deprivation, fatigue, and drug
developed, and elaborated over the course of the in- or alcohol withdrawal.22,23 Individuals who are
terrogation. Most documented false confessions in highly compliant tend to be conflict avoidant, acqui-
recent decades have been directly caused by or have escent, and eager to please others, especially authority
involved promises or threats.5,8 figures.9
The second form of psychological coercion, caus- Highly suggestible or compliant individuals are
ing a suspect to perceive that he has no choice but to not the only ones who are unusually vulnerable to the
comply with the wishes of the interrogator, is not pressures of police interrogation. So are the develop-
specific to any one technique but may be the cumu- mentally disabled or cognitively impaired, juveniles,
lative result of the interrogation methods as a whole. and the mentally ill. The developmentally disabled
If one understands the psychological structure and are more likely to confess falsely for a variety of rea-
logic of contemporary interrogation, it is not difficult sons.24,25 First, because of their subnormal intellec-
to see how it can produce this effect. The custodial tual functioning, low intelligence, short attention
environment and physical confinement are intended span, poor memory, and poor conceptual and com-
to isolate and disempower the suspect. Interrogation munication skills, they do not always understand
is designed to be stressful and unpleasant, and it is statements made to them or the implications of their
more stressful and unpleasant the more intense it answers. They often lack the ability to think in a
becomes and the longer it lasts. Interrogation tech- causal way about the consequences of their actions.
niques are meant to cause the suspect to perceive that Their limited intellectual intelligence translates into
his guilt has been established beyond any conceivable a limited social intelligence as well: they do not al-
doubt, that no one will believe his claims of inno- ways fully comprehend the context or complexity of
cence, and that by continuing to deny the detectives’ certain social interactions or situations, particularly
accusations he will only make his situation (and the adversarial ones, including a police interrogation.
ultimate outcome of the case against him) much They are not, for example, likely to understand that

Volume 37, Number 3, 2009 335

 False Confessions

the police detective who appears to be friendly is vidual, especially if the interrogation is prolonged.
really their adversary or to grasp the long-term con- There have been numerous documented cases of false
sequences of making an incriminating statement. confessions from the developmentally disabled in re-
They are thus highly suggestible and easy to manip- cent years.5
ulate. They also lack self-confidence, possess poor Youth is also a significant risk factor for police-
problem-solving abilities, and have tendencies to induced false confessions.28,29 Many of the develop-
mask or disguise their cognitive deficits and to look mental traits that characterize the developmentally
to others, particularly authority figures, for appropri- disabled may also characterize young children and
ate cues to behavior. It is therefore easy to get them to adolescents. Many juveniles too are highly compli-
agree with and repeat false or misleading statements, ant. They tend to be immature, naively trusting of
even incriminating ones. authority, acquiescent, and eager to please adult fig-
Second, as many researchers have noted, the de- ures. They are thus predisposed to be submissive
velopmentally disabled are eager to please.26 They when questioned by police. Juveniles also tend to be
tend to have a high need for approval and thus are highly suggestible. Like the developmentally dis-
prone to being acquiescent. They have adapted to abled, they are easily pressured, manipulated, or per-
their cognitive disability by learning to submit to and suaded to make false statements, including incrimi-
comply with the demands of others, again especially nating ones. Youth (especially young children) also
those of authority figures.27 Because of their desire to lack the cognitive capacity and judgment to under-
please, they are easily influenced and led to comply in stand the nature or gravity of an interrogation or the
situations of conflict. Some observers refer to this as long-term consequences of their responses to police
“biased responding.”26 The developmentally dis- questions. Like the developmentally disabled, juve-
abled answer affirmatively when they perceive a re- niles also have limited language skills, memory, at-
sponse to be desirable and negatively when they per- tention span, and information-processing abilities
ceive it to be undesirable. They literally tell the compared with normal adults. And juveniles too are
person who is questioning them what they believe less capable of withstanding interpersonal stress and
the questioner wants to hear. A related trait is the thus are more likely to perceive aversive interrogation
cheating-to-lose syndrome. The developmentally as intolerable. All of these traits explain why they are
disabled eagerly assume blame or knowingly provide more vulnerable to coercive interrogation and more
incorrect answers to please, curry favor with, or seek susceptible to making false confessions.
the approval of an authority figure. It is not difficult Finally, people with mental illness are also dispro-
to see how their compliance and submissiveness, es- portionately likely to make false confessions,30 espe-
pecially with figures of authority, can lead the devel- cially in response to police pressure. The mentally ill
opmentally disabled to make false confessions during possess a range of psychiatric symptoms that make
police interrogations. them more likely to agree with, suggest, or confabu-
Third, because of their cognitive disabilities and late false and misleading information and provide it
learned coping behaviors, the developmentally dis- to detectives during interrogations. These symptoms
abled are easily overwhelmed by stress. They simply include faulty reality monitoring, distorted percep-
lack the psychological resources to withstand the tions and beliefs, an inability to distinguish fact from
same level of pressure, distress, and anxiety as men- fantasy, proneness to feelings of guilt, heightened
tally normal individuals.26,27 As a result, they tend to anxiety, mood disturbances, and a lack of self con-
avoid conflict. They may experience even ordinary trol.9,12 In addition, the mentally ill may suffer from
levels of stress, far below that felt in an accusatorial deficits in executive functioning, attention, and
police interrogation, as overwhelming. They are memory; become easily confused; and lack social
therefore less likely to resist the pressures of confron- skills such as assertiveness.30 These traits also increase
tational police questioning and more likely to com- the risk of falsely confessing. While the mentally ill
ply with the demands of their accusers, even if this are likely to make voluntary false confessions, they
means knowingly making a false confession. The may also be easily coerced into making compliant
point at which they are willing to tell a detective what ones. As Salas points out: “Mental illness makes peo-
he wants to hear to escape an aversive interrogation is ple suggestible and susceptible to the slightest form
often far lower than that of a mentally normal indi- of pressure; coercion can take place much more eas-

336 The Journal of the American Academy of Psychiatry and the Law
 Leo

ily, and in situations that a ‘normal’ person might not pressured Lowery to admit to having committed the
find coercive” (Ref. 31, p 264). As a result, “the men- rape because he had recently discovered that his wife
tally ill are especially vulnerable either to giving false was having an affair with another man, and they
confessions or to misunderstanding the context of promised him psychological counseling in lieu of
their confessions, thus making statements against prison, if he admitted to that motive].7 In addition,
their own best interests that an average criminal sus- interrogators encourage the suspect to attribute the
pect would not make” (Ref. 31, p 274). decision to confess to an act of conscience and urge
It is important to emphasize, however, that police him to express remorse about committing the crime.
elicit most false confessions from mentally normal They may provide vivid scene details that appear to
individuals.5,8 For example, Drizin and Leo,5 in a corroborate the suspect’s guilty knowledge and thus
study of 125 proven false confessions, showed that confirm his culpability. Interrogators may also try to
more than 70 percent were given by mentally normal make the admission appear to be voluntary, portray-
individuals (i.e., neither developmentally disabled ing the suspect as the agent of his own confession and
nor mentally ill). themselves merely as its passive recipient.
Interrogators help create the false confession by
The Contamination Error pressuring the suspect to accept a particular account
Psychologically coercive police methods (and how and by suggesting facts of the crime to him, thereby
they interact with an individual’s personality) may contaminating the suspect’s postadmission narrative.
explain how and why a suspect is moved, often pains- Unless he has learned the details of the crime scene
takingly, from denial to admission, but a confession from community gossip or the media, an innocent
is more than an “I did it” statement. It also consists of person would not know either the mundane or the
a subsequent narrative that researchers have referred dramatic details of the crime.8 The innocent sus-
to as the postadmission narrative.8 The narrative pect’s postadmission narrative should therefore be
contextualizes and attempts to explain the “I did it” replete with errors when he responds to questions for
statement and transforms the fledgling admission which the answers cannot easily be guessed by
into a fully formed confession. The postadmission chance, unless, of course, the answers are implied,
narrative makes the story appear, at least on its face, suggested, or explicitly provided to the suspect,
to be a compelling account of the suspect’s guilt. The which, in fact, does occur, whether advertently or
content of and rhetorical force of a suspect’s postad- inadvertently, in many false-confession cases.8,32 If
mission narrative explains, in part, why confessions the entire interrogation is captured on audio or video
are treated as such powerful evidence of guilt and recording, then it may be possible to trace, step by
sometimes lead to the prosecution and conviction of step, how and when the interrogator implied or sug-
the innocent.7 gested the correct answers for the suspect to incorpo-
Police detectives understand the importance of the rate into his postadmission narrative. If, however, the
postadmission phase of interrogation. They use it to entire interrogation is not recorded—and most doc-
influence, shape, and sometimes even script the sus- umented false-confession cases are not—then there
pect’s narrative. The detective’s goal is to elicit a per- may be no objective way to prove that the interroga-
suasive account that successfully incriminates the tor contaminated the suspect’s postadmission
suspect and leads to his conviction. For example, in narrative.
false-confession cases, interrogators have been adept The contamination of the suspect’s postadmission
at inventing, suggesting, or eliciting an account of narrative is thus the third mistake in the trilogy of
the suspect’s motivation; indeed, they often use sce- police errors that, cumulatively, lead to the elicita-
nario-based inducements as a method of attributing a tion and construction of a suspect’s false confession.
motive to the suspect, typically one that minimizes
his culpability, one that the suspect agrees to and The Different Types of False Confession
then repeats, even if it is completely inaccurate. [In In 1985 social psychologists Saul Kassin and Law-
the case of Lowery v. County of Riley, 522 F.3d 1086 rence Wrightsman,33 drawing on case studies and
(10th Cir. 2008), who was accused and ultimately social psychological theories of attitude change, first
convicted of raping an elderly woman and was exon- identified three distinct types of false confession,
erated by DNA evidence many years later, police which they called voluntary (occurring “in the ab-

Volume 37, Number 3, 2009 337

 False Confessions

sence of elicitation”), coerced-compliant (occurring more readily recognize and discount voluntary false
when “the suspect publically professes guilt in re- confessions than those they elicit.9
sponse to extreme methods of interrogation, despite
knowing privately that he or she is truly innocent”), Compliant False Confessions
and coerced-internalized (occurring when the sus- A compliant false confession is one given in re-
pect “actually comes to believe that he or she com- sponse to police coercion, stress, or pressure to
mitted the offense”). Ofshe and Leo20 extended and achieve some instrumental benefit—typically either
modified the initial typology of Kassin and Wrights- to terminate and thus escape from an aversive inter-
man21 to include five distinct types of false confes- rogation process, to take advantage of a perceived
sion: voluntary, stress-compliant, coerced compli- suggestion or promise of leniency, or to avoid an
ant, coerced-persuaded, and noncoerced-persuaded. anticipated harsh punishment.20 Perhaps the most
This classification scheme most accurately captures distinct aspect of compliant false confessions is that
the psychological logic and variation in false confes- they are made knowingly: the suspect admits guilt
sions. For simplicity, researchers have dropped the with the knowledge that he is innocent and that what
various prefixes and simply refer to voluntary, com- he says is false. Compliant false confessions are typi-
pliant, and persuaded false confessions.7 It is impor- cally recanted shortly after the interrogation is over.
tant to understand that there are three conceptually There are several reasons that suspects give com-
distinct psychological processes at work in the pro- pliant false confessions. Kassin and Wrightsman first
duction and elicitation of false confessions. suggested that compliant false confessions arise
“through the coerciveness of the interrogation pro-
Voluntary False Confessions cess” (Ref. 33, p 77). In the pre-modern era of Amer-
Kassin and Wrightsman33 initially defined a vol- ican interrogation, physical coercion, the so-called
untary false confession as one that is offered in the third degree, was the primary cause of such confes-
absence of police interrogation. Voluntary false con- sions. Innocent suspects knowingly falsely confessed
fessions are thus explained by the internal psycholog- to avoid or end physical assaults, torture sessions, and
ical states or needs of the confessor9 or by external the like.38 In the modern era, psychological coercion
pressure brought to bear on the confessor by some- is the primary source of compliant false confessions.
one other than the police.34 Most voluntary false Psychologically oriented interrogation techniques
confessions appear to result from an underlying psy- are just as capable of eliciting compliant false confes-
chological disturbance or psychiatric disorder. As sions as are physical ones. Ofshe and Leo20 have iden-
Kassin and Wrightsman33 point out, individuals vol- tified “classically coercive influence techniques” (i.e.,
unteer false confessions in the absence of police ques- threats and promises, explicit or implied) as the root
tioning for a variety of reasons: a desire for notoriety cause of most compliant false confessions in the
or fame, the need to expiate guilt over imagined or modern era.
real acts, an inability to distinguish between fantasy Coerced-compliant false confessions are the most
and reality, or a pathological need for acceptance or common type of false confession. As Kassin notes,
self-punishment. But voluntary false confessions “the pages of legal history are filled with stories of
need not be rooted in psychological maladies. A per- coerced-compliant confessions” (Ref. 39, p 225).
son may, for example, provide a voluntary false con- Compliant false confessions occur as a result of the
fession out of a desire to aid and protect the real sequenced influence process through which detec-
criminal,35 to provide an alibi for a different crime or tives seek to persuade a suspect that he is indisputably
norm violation,36 or to get revenge on another per- caught and that the most viable way to mitigate his
son.9 High-profile crimes, such as the Lindbergh kid- punishment and escape his otherwise hopeless situa-
napping in the 1930s, the Black Dahlia murder in tion is by confessing.40 As has been well docu-
the 1940s, and the JonBenet Ramsey and Nicole mented, American police use interrogation tech-
Brown Simpson murders in the 1990s, tend to at- niques that are designed, on the one hand, to
tract a large number of voluntary false confessions.37 convince a suspect that he is caught and that it is
Detectives tend to be far more skeptical and less ac- futile for him to deny the crime and, on the other
cepting of voluntary false confessions than of police- hand, techniques that are designed to motivate him
induced false confessions. Put differently, police to perceive that it is in his interest to confess.

338 The Journal of the American Academy of Psychiatry and the Law
 Leo

The most potent psychological inducement is the unfold in three sequential steps. First, the interroga-
suggestion that the suspect will be treated more le- tor causes the suspect to doubt his innocence. This is
niently if he confesses and more punitively if he does typically a by-product of an intense, lengthy, and
not. Unlike most police interrogation techniques, deceptive accusatorial interrogation in which the in-
promises and threats are neither standard nor legal; terrogator repeatedly accuses the suspect of commit-
rather, they are regarded as coercive in both psychol- ting the crime, relentlessly attacks the suspect’s deni-
ogy and law.41 It is not hard to understand why such als (as implausible, illogical, contradicted by the
threats and promises in combination with standard known facts, or simply wrong because of the interro-
interrogation techniques, such as repeated accusa- gator’s alleged superior knowledge or authority) and
tions, attacks on a suspect’s denials, lies about non- repeatedly confronts the suspect with fabricated (but
existent evidence, pressure, and inducements, may allegedly irrefutable) evidence of his guilt. When first
cause a suspect to confess knowingly to a crime he did accused, the innocent suspect thinks that his interro-
not commit. Put simply, the suspect comes to per- gators are genuinely mistaken, and he counters by
ceive that the benefits of confessing (e.g., release from attempting to reason with them and persuade them
custody, mitigated punishment) outweigh the costs of his innocence. At some point, however, the sus-
of denial (e.g., arrest, aggravated punishment). This pect realizes that they are not going to credit his
may be especially true for those suspects who naively assertions of innocence. He may then begin to expe-
believe that the fact of their innocence will, in the rience dissonance because he cannot reconcile the
end, exonerate them.42 obvious contradiction between his knowledge that
Although psychologically coercive threats and he is innocent and his belief that the police are truth-
promises may be the primary sources of compliant fully reporting unmistakable evidence of his guilt.
false confessions, they are not the only ones. Stress20 The suspect offers up the remaining basis for his
and police pressure9 are also causes. Custodial inter- belief in his innocence: that he has no memory of
rogation is inherently stressful, anxiety-provoking, committing the crime.
and unpleasant. The interrogator’s interpersonal To convince the suspect that it is plausible, and
style may also be a source of distress: he may be likely, that he committed the crime, the interrogators
confrontational, insistent, demanding, overbearing, must supply him with a reason that satisfactorily ex-
deceptive, hostile, and manipulative. His accusato- plains how he could have done it without remember-
rial techniques are also designed to induce distress by ing it. This is the second step in the psychological
attacking the suspect’s self-confidence, by not per- process that leads to a persuaded false confession.
mitting him to assert his innocence, and by causing Typically, the interrogator suggests one version or
him to feel powerless and trapped. The interrogation another of a “repressed” memory theory. He or she
may span hours, as often occurs with compliant false may suggest, for example, that the suspect experi-
confessions, weakening a suspect’s resistance, induc- enced an alcohol- or drug-induced blackout, a “dry”
ing fatigue, and heightening suggestibility. blackout, a multiple personality disorder, a momen-
The combined effect of these multiple stressors tary lapse in consciousness, or posttraumatic stress
may overwhelm the suspect’s cognitive capacities disorder, or, perhaps most commonly, that the sus-
such that he confesses simply to terminate what has pect simply repressed his memory of committing the
become an intolerably stressful experience. Facing an crime because it was a traumatic experience for him.
overbearing interrogator who refuses to take no for The suspect can only be persuaded to accept re-
an answer, he may reason that telling the interrogator sponsibility for the crime if he regards one of the
what he wants to hear is the only way to escape.21 interrogators’ explanations for his alleged amnesia as
plausible. Once the suspect is convinced, he comes to
Persuaded False Confessions believe that it is more likely than not that he com-
Persuaded false confessions occur when police in- mitted the crime. The suspect remains in an uncer-
terrogation tactics cause an innocent suspect to tain belief state, because he still has no memory of
doubt his memory and he becomes temporarily per- committing the crime. Despite his lack of memory,
suaded that it is more likely than not that he com- once the suspect is over the line,21 he is ready for the
mitted the crime, despite having no memory of com- third and final step in the making of a persuaded false
mitting it.20 Persuaded false confessions typically confession: the construction of the postadmission

Volume 37, Number 3, 2009 339

 False Confessions

narrative. Once the suspect has accepted responsibil- nates. Regardless, ordinary police interrogation is not
ity for the crime, the interrogator pushes him to sup- strong enough to produce a permanent change in the
ply the details of how and why he did it. The suspect suspect’s beliefs.20
does not know the facts; he is in the paradoxical
situation of believing he committed an act that he
wants to confess to but cannot remember. He may The Consequences of Police-Induced
believe that if he thinks hard enough, searches his False Confessions
mind, or tries to imagine himself committing the Confessions are the most incriminating and per-
crime, he will somehow be able to remember it and suasive evidence of guilt that the state can bring
supply the desired details; but he does not remember against a defendant. False confessions are therefore
the crime. Instead, the suspect either guesses or con- the most incriminating and persuasive false evidence
fabulates about how the crime could have occurred, of guilt that the state can bring against an innocent
repeats the details that the police have suggested to defendant. Former U.S. Supreme Court Justice Wil-
him, knowingly makes up the details, or tries to infer liam Brennan’s observation that “no other class of
them from the interrogators’ suggestions. evidence is so profoundly prejudicial”43 is amply
Usually, the persuaded false confessor’s postad- supported by social science research.44 Confessions
mission narrative is replete with errors. Reasoning exert a strong biasing effect on the perceptions and
from inference rather than actual knowledge, his decision-making of criminal justice officials and lay
confession is given in hypothetical, tentative, and jurors alike because most people assume that a con-
speculative language.9,21 His use of equivocal lan- fession, especially a detailed confession, is, by its very
guage reflects his uncertain belief state, confusion, nature, true. Confession evidence therefore tends to
and lack of memory. Assuming that the suspect is define the case against a defendant, usually overrid-
lying, however, the interrogators sometimes reject his ing any contradictory information or evidence of
speculations and pressure him to use declarative innocence.8
rather than conditional language and to provide the A suspect’s confession sets in motion a seemingly
details of the crime that they continue to believe he irrefutable presumption of guilt among justice offi-
knows. Some persuaded false confessors bend to the cials, the media, the public, and lay jurors.8 This
demands of their interrogators and confess in declar- chain of events in effect leads each part of the system
ative language (e.g., “I did” instead of “I must have to be stacked against the individual who confesses,
done”), even though they lack any knowledge or and as a result he is treated more harshly at every stage
memory of the crime; others continue to use equiv- of the investigative and trial process.45 He is signifi-
ocal, speculative, and uncertain language (“I must cantly more likely to be incarcerated before trial,
have done it,” “I probably did it,” “I guess I did it,” “I charged, pressured to plead guilty, and convicted.
could have done it”), insisting that they still do not Moreover, the presence of a confession creates its
know or remember the details. Ofshe and Leo20 have own set of confirmatory and cross-contaminating bi-
called this the grammar of confabulation. This lan- ases,46,47 leading both officials and jurors to interpret
guage of uncertainty is present in all persuaded false all other case information in the worst possible light
confessions. Usually, the persuaded false confessor for the defendant. For example, a weak and ambigu-
recants either during the interrogation or shortly af- ous eyewitness identification that otherwise may
ter being removed from the interrogation have been quickly dismissed in the absence of a con-
environment. fession is treated instead as corroboration of the con-
Persuaded false confessions appear to occur far less fession’s underlying validity. As the case against an
often than compliant false confessions. They also innocent false confessor moves from one stage to the
tend to occur primarily in high-profile murder cases next in the criminal justice system, it gathers more
and to be the product of unusually lengthy and psy- collective force, and the error becomes increasingly
chologically intense interrogations.7 Once he is re- difficult to reverse.
moved from the interrogation environment and its This chain reaction starts with the police. Once
attendant influences and pressures, the persuaded they obtain a confession, they typically close their
false confessor typically recants his confession.9,20 investigation, clear the case as solved, and make no
Some recant even before the interrogation termi- effort to pursue any exculpatory evidence or other

340 The Journal of the American Academy of Psychiatry and the Law
 Leo

possible leads, even if the confession is internally in- confessions, Leo and Ofshe8,56 found that 73 percent
consistent, contradicted by external evidence, or the of all false confessors whose cases went to trial were
result of coercive interrogation.45 Even when other erroneously convicted; this number went up to 81
case evidence subsequently emerges suggesting or percent in the study of Drizin and Leo5 of 125 false
demonstrating that the suspect’s confession is false, confessions. Taken together, these studies demon-
police almost always continue to believe in the sus- strate that a false confession is a dangerous piece of
pect’s guilt and the underlying accuracy of the con- evidence to put before a judge or jury, because it
fession.5,8 American police interrogators are poorly profoundly biases their evaluation of the case in favor
trained about the risks of psychological interrogation of conviction, so much so that they may allow it to
and the phenomenon of police-induced false confes- outweigh even strong evidence of a suspect’s factual
sion,8,40 and, like most people, they tend to assume innocence.8 The Leo and Ofshe8,56 and Drizin and
that virtually all confessions are true and thus assume Leo5 studies show that real-world jurors simply fail to
that virtually all who confess are guilty.7 discount false-confession evidence appropriately,
The presumption of guilt and the tendency to even when the defendant’s uncorroborated confes-
treat more harshly those who confess extends to pros- sion was elicited by coercive methods and the other
ecutors, defense attorneys, and judges as well. Like case evidence strongly supports his innocence. Thus
police, prosecutors rarely consider the possibility that the false-confession evidence is highly, if not inher-
an entirely innocent suspect has falsely confessed; ently, prejudicial to the fate of any innocent defen-
some prosecutors are so skeptical of the idea of po- dant in the American criminal justice system. As
lice-induced false confession that they stubbornly Welsh White notes, “the system does not have safe-
refuse to admit that it occurred, even after DNA guards that will prevent the jury from giving dispro-
evidence has unequivocally established the defen- portionate weight to such confessions” (Ref. 41, p
dant’s innocence.48 Once a suspect has confessed, 155).
prosecutors tend to charge him with the highest The findings from these field studies of aggregated
number and types of offenses49 and set his bail at a false-confession cases are consistent with those from
higher amount50 (especially in serious or high profile experiments and public opinion surveys. They all
cases), and they are far less likely to initiate or accept converge on the same conclusion: that, as the U.S.
a plea bargain to a reduced charge.21 The confession Supreme Court stated in the case of Arizona v. Ful-
becomes the centerpiece of the prosecution’s case minante, “a confession is like no other evidence.”57 It
against the defendant. Even defense attorneys treat is “uniquely potent”58 in its ability to bias the trier of
suspects who confess more harshly, often pressuring fact in favor of the prosecution, overwrite contradic-
them to accept a guilty plea to a lesser charge to avoid tory or exculpatory case evidence, and lead to the
the higher sentence that will inevitably follow from a wrongful conviction of the innocent.8 Researchers
jury conviction.51,52 As the California Supreme have demonstrated that mock jurors find confession
Court has noted, “the confession operates as a kind of evidence more incriminating than any other type of
evidentiary bombshell which shatters the defense.”53 evidence.44,58 Kassin and Sukel55 found that confes-
American judges too tend to presume that a defen- sions powerfully increased the conviction rate even
dant who has confessed is guilty and, accordingly, when mock jurors viewed the confession as coerced,
treat him more punitively. Conditioned to disbelieve even when they were instructed to disregard the con-
defendants’ claims of innocence or police miscon- fession as inadmissible, and even when they reported
duct, judges rarely suppress confessions, even highly that it had no influence on their verdict. Confessions,
questionable ones.54 especially detailed confessions, are devastating to a
If the defendant’s case goes to trial, the jury treats defendant’s case because, as Welsh White notes, “ju-
the confession as more probative of the defendant’s rors are often unwilling to believe that anyone would
guilt than any other type of evidence (short of a vid- confess to a crime that they did not commit” (Ref.
eotape of the suspect committing the crime), espe- 59, p 134).
cially if, as in virtually all high profile cases, the con-
fession receives pretrial publicity.8,44,55 False Conclusions
confessions are highly likely to lead to the wrongful As this article has shown, empirical researchers
conviction of the innocent. In their study of 60 false have documented and analyzed how and why con-

Volume 37, Number 3, 2009 341

 False Confessions

temporary methods of psychological interrogation the increasing documentation of false confessions

can, and sometimes do, lead innocent individuals to and the interrogative influences that promote them
confess falsely to serious crimes. The consequences of and ultimately lead to the conviction of the innocent.
these false confessions are disastrous for innocent in-
dividuals who are wrongfully convicted and incarcer- Acknowledgment
ated. As a result, empirical researchers have also sug- I thank Alan Law for research assistance.
gested ways to minimize both the number of false
confessions that police elicit and the number of false References
confessions that, once elicited, lead to wrongful con- 1. Leo R: Re-thinking the study of miscarriages of justice: developing
a criminology of wrongful conviction. J Contemp Crim Just 21:
victions. Mandatory electronic recording of police 201–23, 2005
interrogations in their entirety is the single most im- 2. Gross S, Jacoby K, Matheson D, et al: Exonerations in the United
portant policy reform available because it creates an States, 1989 through 2003. J Crim Law Criminol 95:523–53,
2005
objective, comprehensive, and reviewable record of 3. Borchard EM: Convicting the Innocent: Errors of Criminal Jus-
the interrogation that all parties—police, prosecu- tice. New Haven: Yale University Press, 1932
tors, defense attorneys, judges, juries, and even the 4. Baumgartner FR, DeBoef SL, Boydstun AE: The Decline of the
Death Penalty and the Discovery of Innocence. Cambridge, UK:
public in high-profile cases— can review. Although Cambridge University Press, 2007
10 states (Alaska, Minnesota, Illinois, Maine, New 5. Drizin S, Leo R: The problem of false confessions in the post-
Mexico, Wisconsin, New Jersey, North Carolina, DNA world. NC L Rev 82:891–1007, 2004
6. Garrett B: Judging innocence. Colum L Rev 108:55–142, 2008
Maryland, and Nebraska) and the District of Colum- 7. Leo R: Police Interrogation and American Justice. Harvard Uni-
bia now require that police record interrogations in versity Press, 2008
their entirety in some or all criminal cases, most po- 8. Leo R, Ofshe R: The consequences of false confessions: depriva-
lice departments, as well as the FBI, still do not tions of liberty and miscarriages of justice in the age of psycholog-
ical interrogation. J Crim Law Criminol 88:429 –96, 1998
record interrogations, and there remains resistance to 9. Gudjonsson GH: The Psychology of Interrogations and Confes-
the idea in many quarters of law enforcement.60 Re- sions: A Handbook. New York: Wiley, 2003
searchers have proposed other reforms as well, in- 10. Leo R: Questioning the relevance of Miranda in the twenty-first
century. Mich L Rev 99:1000 –29, 2001
cluding improved police training about false confes- 11. Chojnacki DE, Cicchini MD, White LT: An empirical basis for
sions, pretrial reliability hearings to exclude false- the admission of expert testimony on false confessions. Ariz St L J
confession evidence, putting time limits on 40:1, 2008
12. Kassin S, Gudjonsson G: The psychology of confessions: a review
interrogations, prohibiting certain interrogation of the literature and issues. Psychol Sci Public Int 5:35– 67, 2004
techniques, greater provision of expert witness testi- 13. Zimbardo P: Coercion and compliance: the psychology of police
mony and cautionary jury instructions at trial, and confessions, in The Triple Revolution. Edited by Perruci C,
Pilisuk M. Boston: Little, Brown, 1971, pp 492–508
providing additional safeguards for vulnerable pop- 14. Davis D, Leo R: Strategies for preventing false confessions and
ulations such as the developmentally disabled and their consequences, in Practical Psychology for Forensic Investi-
juveniles.7 gations and Prosecutions. Edited by Kebbell M, Davies G. New
Such reforms, however, are likely to occur slowly York: John Wiley & Sons, 2006, pp 121– 49
15. Bond CF, DePaulo BM: Accuracy of deception judgments. Pers
in the United States. Great Britain has adopted sev- Soc Psychol Rev 10:214 –34, 2006
eral reforms, based on growing documentation and 16. Hartwig M, Granhag PA, Stromwall L, et al: Police officers’ lie
awareness of the problem of false confessions.9 Amer- detection accuracy: interrogating freely vs. observing video. Police
Q 7:429 –56, 2004
ican law enforcement, however, remains steeped in 17. Kassin S, Fong CT: “I’m innocent!” Effects of training on judg-
the use of investigative methods and interrogation ments of truth and deception in the interrogation room. Law
techniques that continue to cause the three errors Hum Behav 23:499 –516, 1999
18. Davis D, Follette W: Rethinking probative value of evidence: base
that produce false confessions, and the American rates, intuitive profiling and the postdiction of behavior. Law
public continues to believe in the myth of psycholog- Hum Behav 26:133–58, 2002
ical interrogation. Until the misconception that in- 19. Gross S: The risks of death: why erroneous convictions are com-
mon in capital cases. Buff L Rev 44:469 –500, 1996
nocent suspects do not confess in response to psycho- 20. Ofshe R, Leo R: The social psychology of police interrogation: the
logical interrogation is dispelled, police detectives theory and classification of true and false confessions. Stud Law
will continue to elicit false confessions that lead to Politics Soc 16:189 –251, 1997
wrongful convictions. As a result, social scientists and 21. Ofshe R, Leo R: The decision to confess falsely: rational choice
and irrational action. Denv U L Rev 74:979 –1122, 1997
mental health professionals must continue to con- 22. Blagrove M: Effects of length of sleep deprivation on interrogative
duct empirical research and educate the public about suggestibility. J Exp Psychol 2:48 –59, 1996

342 The Journal of the American Academy of Psychiatry and the Law
 Leo

23. Harrison Y, Horne J: The impact of sleep deprivation on decision 40. Davis D, O’Donahue W: The road to perdition: extreme influ-
making: a review. J Exp Psychol 6:236 – 49, 2000 ence tactics in the interrogation room, in Handbook of Forensic
24. Cloud M, Shepherd G, Barkoff A, et al: Words without meaning: Psychology. Edited by O’Donahue W. San Diego: Academic
the constitution, confessions and developmentally disabled sus- Press, 2004, pp 897–996
pects. U Chi L Rev 69:495– 624, 2002 41. White WS: Miranda’s Waning Protections: Police Interrogation
25. Conley R, Luckasson R, Bouthilet G: The Criminal Justice Sys- Practices After Dickerson. Ann Arbor, MI: University of Michi-
tem and Mental Retardation: Defendants and Victims. Baltimore: gan Press, 2001
Paul H. Brookes, 1992 42. Kassin S: On the psychology of confessions: does innocence put
26. Ellis J, Luckasson R: Developmentally disabled defendants. Geo innocents at risk? Am Psychol 60:215–28, 2005
Wash L Rev 53:414 –93, 1985 43. Colorado v. Connelly, 479 U.S. 157, 182 (1986)
27. Gudjonsson G, Clare I, Rutter S, et al: Persons at risk during 44. Miller GR, Boster FJ: Three images of the trial: their implications
interviews in police custody: the identification of vulnerabilities. for psychological research, in Psychology in the Legal Process.
Royal Commission on Criminal Justice Research Study No. 12.
Edited by Sales B. New York: Pocket Books, 1977, pp 19 –38
London: Her Majesty’s Stationery Office, 1993
45. Leo R: Inside the interrogation room. J Crim Law Criminol 86:
28. Owen-Kostelnik J, Reppucci N, Meyer J: Testimony and interro-
gation of minors: assumptions about maturity and morality. Am 266 –303, 1996
Psychol 4:286 –304, 2006 46. Castelle G, Loftus E: Misinformation and wrongful convictions,
29. Drizin S, Colgan B: Tales from the juvenile confession front: a in Wrongly Convicted: Perspectives on Failed Justice. Edited by
guide to how standard police interrogation tactics can produce Westervelt SD, Humphrey JA. Newark: Rutgers University Press,
coerced and false confessions from juvenile suspects, in Interroga- 2001, pp 17–35
tions, Confessions, and Entrapment. Edited by Lassiter GD. New 47. Findley K, Scott M: The multiple dimensions of tunnel vision in
York: Kluwer Academic, 2004, pp 127– 62 criminal cases. Wis L Rev 291–398, 2006
30. Redlich A: Mental illness, police interrogations, and the potential 48. Medwed D: The zeal deal: prosecutorial resistance to post-convic-
for false confession. Law Psychiatry 55:19 –21, 2004 tion claims of innocence. B U L Rev 84:125– 83, 2004
31. Salas C: The case for excluding confessions of the mentally ill. Yale 49. Cassell P, Hayman B: Police interrogation in the 1990s: an em-
J L & Human 16:243–75, 2004 pirical study of the effects of Miranda. UCLA L Rev 43:839 –931,
32. Leo R, Drizin S, Neufeld P, et al: Bringing reliability back: false 1996
confessions and legal safeguards in the twenty-first century. Wis L 50. Walker S: Popular Justice: A History of American Criminal Jus-
Rev 2:479 –539, 2006 tice (ed 2). Oxford, UK: Oxford University Press, 1998
33. Kassin S, Wrightsman L: Confession evidence, in The Psychology 51. Nardulli P, Eisenstein J, Fleming R: The tenor of justice: criminal
of Evidence and Trial Procedure. Edited by Kassin S, Wrightsman courts and the guilty plea process. Urbana, IL: University of Illi-
L. Beverly Hills: Sage Publications, 1985, pp 67–94 nois Press, 1988
34. McCann J: A conceptual framework for identifying various types 52. Wald M, Ayres R, Hess DW, et al: Interrogations in New Haven:
of confessions. Behav Sci Law 16:441–53, 1998 the impact of Miranda. Yale L J 76:1519 – 648, 1967
35. Wagenaar W, van Koppen P, Crombag H: Anchored Narratives: 53. California v. Cahill, 5 Cal.4th 497 (Cal. 1993).
The Psychology of Criminal Evidence. New York: St. Martin’s 54. Givelber D: Punishing protestations of innocence: denying re-
Press, 1993 sponsibility and its consequences. Am Crim L Rev 37:1363– 408,
36. Radelet M, Bedau H, Putnam C: In Spite of Innocence: Errone-
2000
ous Convictions in Capital Cases. Boston: Northeastern Univer-
55. Kassin S, Sukel H: Coerced confessions and the jury. Law Hum
sity Press, 1992
37. Corwin M: False confessions and tips still flow in Simpson case, Behav 21:27– 46, 1997
crime: Such calls have been common since Lindbergh kidnapping. 56. Leo R, Ofshe R: The truth about false confessions and advocacy
About 500 confessed to Black Dahlia killing. Los Angeles Times, scholarship. Crim Law Bull 37:293–370, 2001
March 25, 1996, p A1. See also, Voluntary false confessions: a 57. Arizona v. Fulminante, 499 U.S. 279, 296 (1991)
neglected area in criminal administration. Ind L J 28:374 –392, 58. Kassin S, Neumann K: On the power of confession evidence: an
1953 experimental test of the fundamental difference hypothesis. Law
38. Leo R: The third degree and the origins of psychological police Hum Behav 21:460 – 84, 1997
interrogation in the United States, in Interrogations, Confessions, 59. White WS: False confessions and the constitution: safeguards
and Entrapment. Edited by Lassiter GD. New York: Kluwer Ac- against untrustworthy evidence. Harv CR-CL L Rev 32:105–57,
ademic, 2004, pp 37– 84 1997
39. Kassin S: The psychology of confession evidence. Am Psychol 60. Leo R, Richman KD: Mandate the electronic recording of police
52:221–33, 1997 interrogations. Crime Public Policy 6:791– 8, 1997

Volume 37, Number 3, 2009 343