Vol. 80, No. 5, May 2007
ancy DePriest was raped and murdered in Austin, Texas, in 1988, as she opened the Pizza Hut restaurant she managed.1 The brutal attack shocked the surrounding community. There were no obvious suspects and few leads, but the public quickly began pressuring police to solve the crime.
Christopher Ochoa lived in Austin, had no prior criminal record, and did not know DePriest. Several weeks after the murder, Ochoa and a friend, Richard Danziger, went to the Pizza Hut where DePriest was murdered. Employees overheard the two men asking questions about the crime and alerted police, who brought Ochoa and Danziger in for questioning.
During two 12-hour interrogation sessions, detectives repeatedly told Ochoa that they knew he had murdered DePriest. They cut off and ignored his repeated denials. They lied to him, saying they had strong evidence against him. They told him that he would face execution if he did not confess. Ochoa eventually broke down. He signed a lengthy confession containing details that only law enforcement officers and the killer would have known.
Ochoa told his court-appointed attorney that his confession was false and that he was innocent. But his attorney did not believe him and conducted no investigation. Instead, the attorney successfully pressured Ochoa to plead guilty to avoid execution. As part of his plea deal, Ochoa testified against Danziger. Both men were convicted and sentenced to life in prison.
A decade later, a convicted serial rapist named Achim Marino contacted authorities from his prison cell and stated that he alone had killed DePriest. Marino insisted that Ochoa and Danziger were innocent. But Marino was ignored because Ochoa's confession seemed airtight. Ochoa's new attorneys, despite being told by the trial attorney that there was "not a chance" that Ochoa was innocent, arranged for DNA testing on semen found in the rape kit. The testing indisputably identified Marino as DePriest's killer. Ochoa and Danziger were released. Ochoa later became a lawyer; Danziger, however, sustained brain damage from a prison fight.
Ochoa's case is not the only one of its kind. Recent exonerations, many through DNA testing, have revealed that false confessions occur more often than one might expect and play a substantial role in wrongful convictions. Of the first 177 DNA exonerations, 41 (23 percent) involved cases in which the defendant gave an interrogation-induced false confession.2 Researchers have documented many other proven false confessions in non-DNA cases.3 In Illinois, of 42 people known to have been wrongly convicted of homicide since 1970, 25 (approximately 60 percent) falsely confessed.4 These cases, of course, are only part of the picture, because they describe only discovered false confessions. The total rate of true and false confessions is unknown.
At the same time, common sense and experience suggest that the vast majority of confessions are true. Police interrogation, a long-time staple of our criminal justice system, is a necessary element of crime-fighting. Because not all cases contain physical evidence, and because of resource limitations and public pressure for speedy apprehension of guilty people, interrogation is, and will remain, a necessity.
To help ensure the reliability of interrogation evidence, this article describes research about the causes of false confessions and possible ways to prevent false confession-based wrongful convictions. There is not unanimous agreement on these topics among researchers or practitioners, but research provides a useful starting point. Although false confession-based wrongful convictions often begin with police interrogation, they involve a breakdown of the entire justice system and include failures by prosecutors, defense attorneys, and judges. Everyone in the system has a role to play in addressing false confessions.
Categories of False Confessions
Researchers have isolated three kinds of false confessions5:
1) Voluntary false confessions sometimes occur in highly publicized cases, such as the JonBenet Ramsey case, in which a man confessed without any interrogation but was released when the confession proved unreliable.
2) Coerced-compliant false confessions occur when a suspect chooses to falsely confess in order to end a stressful interrogation or to receive some benefit, such as a reduced charge.
3) Coerced-internalized false confessions occur when a suspect under interrogation comes to falsely believe that he or she committed the crime but cannot remember doing so; often such confessions occur because of the influence of drugs or mental illness.
Causes of False Confessions
The Rational Choice Theory. To understand the two most common kinds of false confessions, coerced-compliant and coerced-internalized, researchers developed the rational choice theory, which posits that some suspects falsely confess because interrogation convinces them that confessing is the most rational choice.6 This can occur when an innocent suspect believes that the police have strong evidence and that confessing is the only way to avoid severe punishment. Or, it can occur if an innocent suspect, wanting to end a stressful interrogation, believes that it will be impossible to shake the interrogator's belief in guilt but that proof of innocence will emerge later.
The Role of Confrontational Interrogation Techniques. Based on the rational choice theory, researchers have argued that confrontational interrogation techniques - the most prominent being the Reid technique - risk eliciting false confessions.7 The Reid technique is a two-stage process: 1) a nonaccusatory interview, and 2) an accusatory interrogation.8 During the nonaccusatory interview, the interviewer asks open-ended questions and evaluates the suspect's truthfulness using behavior symptom analysis, which means looking for behaviors such as lack of eye contact, shifting posture, and fidgeting, to determine whether the suspect is being deceptive.9
If the interviewer believes the suspect is guilty after the nonaccusatory interview, the interviewer then conducts an accusatory interrogation, which is dominated by the interrogator, with the goal of getting the suspect to incriminate himself or herself. An accusatory interrogation typically begins with a technique sometimes called maximization, which is an elaborate process designed to break down the suspect and make his or her situation seem hopeless by suggesting that conviction is inevitable. The interrogator repeatedly states the absolute certainty of the suspect's guilt, cutting off all denials of guilt, and confronts the suspect with evidence, whether real or fabricated.
Building on this hopelessness, the interrogator then employs strategies designed to convince the suspect that, given the irrefutable evidence, confessing is the best option available. These strategies are known as minimization, because they consist of either minimizing the seriousness of the offense or convincing the suspect that confessing will minimize the consequences of being convicted.
Behavior Symptom Analysis. Researchers have criticized behavior symptom analysis, the process of human lie detection that directs interrogators to judge truthfulness by evaluating nonverbal behavior. Researchers have highlighted laboratory research suggesting that people cannot detect deception at better-than-chance rates10 and have suggested that behavior symptom analysis may lead interrogators astray, causing them to misinterpret innocent behaviors as deceptive and to interrogate innocent suspects.
Proponents of the Reid technique, however, have argued that the research criticizing behavior symptom analysis is not applicable to the real world of police interrogation because the research setting does not replicate real-life conditions. These commentators argue that the laboratory subjects lacked the motivation to lie convincingly, and that the laboratory interrogators had less information about the subjects than real interrogators would have about real suspects.
Deception. Researchers also have criticized the use of deception. Although not barred by courts,12 deception is a recurring factor in interrogation-induced false confessions: false evidence ploys - in which the interrogator lied about the evidence - were used in many of the false confession cases.13 False evidence ploys may increase the risk of false confessions by leading suspects to believe that the police have strong evidence, that it will be impossible to prove innocence, and that confessing is therefore the only way to avoid harsh punishment.14
Duration of Interrogation. Researchers also argue that lengthy interrogations increase the risk of false confessions. Research suggests that most interrogations are short, with suspects providing information in approximately 30 minutes.15 But a high percentage of proven false confessions occurred after unusually long interrogations. One study of proven false confessions found that, of those cases in which the length of the interrogation was recorded, the average length was 16.3 hours.16 A long interrogation causes fatigue and stress, exacerbates hopelessness, and may contribute to a false confession from a suspect who wants the interrogation to end.
Suspects' Characteristics. Research suggests that the risk of false confessions is greater with certain kinds of suspects, such as juveniles, mentally ill people, or suspects with drug or alcohol addictions.17 In the same sample of 125 proven false confessions, approximately one-third (40) of the persons who made confessions were juveniles, and 22 of these 40 were under age 15.18
Alternatives to Confrontational Interrogation Techniques. Law enforcement entities have begun exploring alternatives to confrontational techniques. St. Paul, Minn., police commander Neil Nelson pioneered a technique, tailored to electronic recording, in which the interrogator does not cut off denials, but instead encourages suspects to provide as many details as possible so that the interrogator can elicit provable lies. Nelson tailored his technique to electronic recording, in part because he found that jurors viewing recorded interrogations sometimes disliked interrogators who used confrontational techniques. Nelson also found that, when dealing with hardened criminals, confrontational techniques rarely succeeded in eliciting confessions. A technique similar to Nelson's, called investigative interviewing,19 was implemented in Britain after authorities concluded that confrontational techniques are unnecessary to eliciting incriminating statements and actually inhibit interrogators from obtaining useful information.20
Corroboration. Experienced investigators have long emphasized the importance of corroborating a confession's details through independent investigation. The proven false confession cases underscore this point. In some cases, such as Christopher Ochoa's, the suspect signed a confession written by police, in which the details came not from the suspect but from police knowledge. In other cases, such as the infamous Central Park jogger case, the five confessions did not match each other in many details and contained significant errors compared to the known facts.
To avoid these problems, researchers and experienced investigators have recommended eliciting a detailed post-admission narrative from a suspect who has confessed and then evaluating whether the narrative fits with known facts.
Searching for independent corroboration also helps avoid tunnel vision, a set of cognitive tendencies that lead people to seek out and filter information in a way that confirms pre-existing beliefs.21 In the context of confessions, it is natural for an investigator who has elicited a confession to believe that the confessor is guilty. But investigators who search for independent corroboration will be more likely to exhaust independent sources of evidence and therefore more likely to discover those rare false confessions.
Existing Safeguards against False Confessions
Miranda and Voluntariness. A confession is inadmissible if taken in violation of Miranda, or if the confession was involuntary because the suspect's will was overwhelmed by interrogators.22 Although the Miranda and voluntariness rules were designed partly to ensure the reliability of confessions, both rules have been ineffective in preventing false confessions. First, most people waive their Miranda rights and participate in interrogation.23 Some people, especially mentally ill people, drug or alcohol addicts, and juveniles, fail to understand Miranda warnings.24 Other innocent people waive their Miranda rights because they believe they have nothing to hide and will be able to prove their innocence.
The voluntariness rule also has been ineffective at preventing false confessions, primarily because courts almost never find involuntariness. Indeed, under the rational choice model, most false confessions result from voluntary choices made by suspects who think their situations will improve if they confess.
Electronic Recording. Police in Wisconsin are now required to electronically record custodial interrogations of juveniles in all cases and custodial interrogations of adults in felony cases.25
Electronic recording should reduce the chances of suspects falsely confessing and should help juries reject false confessions, but it may not completely solve the problem. First, because the remedy for failing to record in adult cases is a jury instruction, not suppression of the confession, unrecorded confessions may still reach juries. Second, as occurred in the Central Park jogger case,26 jurors sometimes will convict a defendant based on a false confession even if the confession is recorded.
Preventing False Confessions from Becoming Wrongful Convictions
Police interrogation is only the first step in a false confession-based wrongful conviction. Much more must happen - through the actions of many other justice system participants - before a false confession becomes a wrongful conviction.
Prosecutors. Before making a charging decision in a confession case, prosecutors evaluate the confession's trustworthiness. There is no foolproof method for doing so, but prosecutors should begin with an awareness that false confessions, counterintuitive though they are, do sometimes occur.
Apart from acknowledging the possibility of false confessions, prosecutors should be aware of factors that may distinguish true confessions from false confessions. First, the characteristics of the confessor (such as whether the suspect is a juvenile, is mentally ill, or has a drug/alcohol addiction) may affect the likelihood of a false confession.
Second, prosecutors should consider the events that led to the confession, such as what interrogation techniques preceded the confession. Very long interrogations, or interrogations in which the police used false evidence ploys, may raise the likelihood of a false confession.
Finally, prosecutors should consider how the confession fits with the other evidence. Prosecutors should evaluate whether the confession: 1) contains details that only the true perpetrator would know, 2) directs police to evidence of which they previously were unaware, or 3) contains errors that do not match the known facts.27
Defense Attorneys. Poor representation by defense attorneys is a critical factor in wrongful convictions. Like prosecutors, defense attorneys must acknowledge the possibility that a client has falsely confessed. Unlike other justice system actors, defense attorneys are not free to disregard a client's claim of innocence. When the client insists that his or her confession is false, the defense attorney should treat it as such and present a zealous false confession defense. This includes becoming familiar with the literature on false confessions. It also may include calling an expert or requesting a jury instruction about factors relevant to false confessions.
Judges. Like prosecutors and defense attorneys, judges also are sometimes required to evaluate the reliability of confessions. This occurs when judges consider: 1) challenges under the corroboration rule,28 2) harmless error or prejudice under the ineffective-assistance-of-counsel test in confession cases, 3) motions to present expert testimony or a jury instruction on false confessions, and 4) motions to suppress a confession under Wis. Stat. section 904.03.29 Thus, like prosecutors and defense attorneys, judges must recognize the possibility (albeit rare) of false confessions and the above-described risk factors.
Although false confessions are counterintuitive, DNA exonerations and social science research have suggested that they do occur and sometimes result from rational choices made during interrogation. Preventing false confession-based wrongful convictions requires criminal justice participants to accept the possibility of false confessions, understand their risk factors, and consider factors useful to evaluating whether a confession is true or false.