June 01, 2011

False Confessions

Confessions are powerful in the courtroom, but not all confessions are genuine. Sometimes details that make a confession seem believable are unintentionally supplied by the police.

Brandon L. Garrett

Download a printable PDF of this article (membership required).

In January 1984, a 16-year-old girl disappeared in Detroit, Michigan. Neighbors searched and found her body in an abandoned parking garage. She had been brutally raped and strangled. Forensic analysts examined swabs from the victim and a green bottle used to assault her. They detected sperm, but no further analysis was done. Months passed, and the crime remained unsolved.

Then a young man named Eddie Joe Lloyd offered to help. He was involuntarily committed in the Detroit Psychiatric Institute, a mental hospital, with a preliminary diagnosis of bipolar affective disorder. He had written letters to police about several unsolved murders, and he said a man had told him about the death of this 16-year-old girl. He mentioned a green bottle. His mental illness included “delusions that he had a special ability to solve crimes.” Jeremy W. Peters, “Wrongful Conviction Prompts Detroit Police to Videotape Certain Interrogations,” N.Y. Times, Apr. 11, 2006. Despite his manifest mental illness, the hospital allowed Lloyd to be interviewed by police, and they did interview him several times.

Police said that Lloyd ultimately did more than try to help them solve a case. They said that when they spoke to him, he suddenly said, “I want to confess to the killing.” He described the killing in detail. Police claimed he had been “anxious to give a formal statement.” He finally signed a written statement and gave a tape-recorded statement. He was arrested and charged with murder. At his criminal trial in 1985, the jury heard the taped statement. There was no other evidence of guilt.

Yet, the details Lloyd gave in the confession were damning. Lloyd apparently described how the victim had been sodomized with a green bottle. He described the strangulation of the victim using a ligature and showed how he grabbed her throat. His account was consistent with the medical examiner’s conclusions of how she had been killed. He told police what clothes the victim wore, including a sweater, gold-colored earrings, and, most specific, “her clean Gloria Vanderbilt jeans.” He did more. He described the layout of the crime scene, with a diagram of it, and he described leaving the victim’s underwear on a tree, which the prosecution termed a sort of “calling card.”

There were a few inconsistencies. For example, he incorrectly identified the location of the body. But the other details were so numerous and so specific that the evidence of guilt was overwhelming.

At the trial, the psychiatrist at the hospital said that although Lloyd was on antipsychotic medication, he was able to understand what he was doing and the consequences when he confessed to police.

The police then testified about the power of those confession statements. The lead officer explained the “standard policy” of keeping key “confidential” crime scene information “until time of trial” because such information can be used to test leads. The officer explained that Lloyd had from the outset told them “confidential information known only to myself and to the members of the squad who made the scene that morning at the garage. The information was in regards to a bottle which had been inserted in a particular area of the victim. This was known only to us.”

The officer underscored that it was Lloyd who “brought up the bottle and this time he brought up the color of the bottle which hadn’t been mentioned.” In addition, “he physically showed how she was laying on her side, which is exactly how she was in the garage in the photographs we have with the knot around her neck.” The officer emphasized the specific details about the victim’s jeans, saying, “He brought up another important fact regarding her Gloria Vanderbilt jeans. . . . He said recently washed, her clean Gloria Vanderbilt jeans. He described the color of the stitching, the independent patch they had in the back of the pants.”

The state put on more than a dozen witnesses, including experts. In the closing arguments to the jury, the prosecutor emphasized that it was suspicious that Lloyd had come forward to talk to the police in the first place. Most convincing, “how would Mr. Lloyd remember all of these details?” The prosecutor explained how the crime scene “was secured, no one else was allowed in.” Yet, Lloyd had supposedly confessed to crime details that “were never disclosed through the media.” The prosecutor told the jury that the only ones who knew the details of the crime “were the homicide investigators and the killer.” The prosecutor emphasized that the taped confession “was shocking and graphic in the detail that he gave. Keep in mind that this was not a taped statement in which the police officers were spoon-feeding information to Mr. Lloyd where he was just parroting back answers like yes or no.”

A Tepid Defense

Lloyd’s counsel offered only a tepid defense. The original, court-appointed defense attorney withdrew eight days before trial, having done little work on the case. A new lawyer, also court-appointed, never met with his predecessor. This lawyer called no witnesses; he originally planned to call Lloyd but chose not to when the judge ruled that Lloyd’s prior convictions would be admissible. Nor did he assert his client’s innocence in his closing arguments. Instead he told the jury that the “so-called confession” left him “bewildered,” concluding, “I ask that you consider this case and you come out with a decision after reviewing everything which I hope you’ll do that is equitable to my client and to the People.” 3 Transcript of Record at 37, People v. Lloyd, No. 85-00376 (Mich. Rec. Ct. May 2, 1985).

The jury deliberated for 40 minutes before convicting Lloyd of first-degree murder. The judge thanked them, saying, “I don’t think that there could have been any other verdict that you could have brought in considering the evidence.”

Lloyd was sentenced to life in prison without any possibility of parole. The judge complained that since Michigan had abolished the death penalty, he could not hang Lloyd, complaining that “I cannot impose the sentence that the facts call for.”

For years, Lloyd pursued appeals and postconviction remedies with no success. He repeatedly raised a series of claims, including that his confession statements should not have been admitted at trial. He made no claim to relief on the basis of innocence but, in a pro se habeas petition, told the court that he was “‘ABSOLUTELY’, ‘TOTALLY’ and ‘COMPLETELY INNOCENT.’ (Lloyd’s emphasis).

All of the courts denied relief. The federal district court found that the evidence that his confession was voluntary was “overwhelming,” despite the fact that Lloyd was in a mental hospital and had been prescribed a “heavy tranquilizer” during the time he spoke with police. The court emphasized that, according to the police, Lloyd had repeated details that were “known only to police.”

The green bottle then entered the picture again. In 1995, Lloyd contacted the Innocence Project at Cardozo Law School. Founded by Barry Scheck and Peter Neufeld, the pioneering Innocence Project has successfully used DNA testing to exonerate individuals erroneously convicted of crimes. The Wayne County District Attorney’s Office and the Detroit Police Department initially “refused to cooperate in finding any biological evidence.” A court order was required, and even then it took years to locate and test the crime scene evidence. Finally, in 2002, DNA testing conducted by two labs on multiple pieces of evidence, including swabs from the victim and from the green bottle, all confirmed that the DNA found at the crime scene did not match Lloyd’s DNA.

After spending 17 years in prison, Lloyd was exonerated.

Lloyd’s case, as remarkable as it may seem, is typical of false confessions elicited by the police from those who were eventually exonerated by virtue of DNA evidence. Forty of the 250 DNA exoneration cases (16 percent) involved a false confession.

History of False Confessions

For decades, many doubted that a suspect would ever falsely confess. For example, Professor John Henry Wigmore wrote in his 1923 A Treatise on the Anglo-American System of Evidence in Trials at Common Law that false confessions were “scarcely conceivable” and “of the rarest occurrence.” Of course, we know that police torture can coerce a false confession, but suspects now routinely confess without illegal use of the “third degree.” After all, the U.S. Supreme Court now only permits an interrogation to be introduced in court if it was “voluntary” under the totality of the circumstances. Police must give suspects the famous Miranda warnings to let them know they have the right to a lawyer and the right to remain silent. When a suspect voluntarily decides to talk, a confession is powerful, nearly uncontestable evidence of guilt.

This traditional understanding of confessions has changed dramatically in recent years, as DNA testing has exonerated people who falsely confessed. While we do not know how often false confessions occur, there is an awareness that innocent people do falsely confess. Many who do so are juveniles or mentally disabled. Such individuals are vulnerable to police pressure. Others without such vulnerabilities may nonetheless eventually succumb to police interrogation tactics. Studies now examine the complex psychological techniques police use that can cause such false confessions.

False confessions also pose a puzzle: How can innocent people confess in any detail to crimes they did not commit? An innocent person might eventually cave in to police pressure and finally say, “I did it.” Just a bare admission of guilt is not very convincing though, and an innocent person could not elaborate on details about how the crime was committed if those details were known only to the guilty individual and to the police.

A false confession may become detailed, though, if police, intentionally or not, prompt the suspect on how the crime happened. A complete and accurate description of the crime makes for a very compelling confession. Although police can use a range of psychologically coercive and even deceptive tactics, they are trained never to feed details to a suspect, thereby contaminating a confession. The leading manual on police interrogations, originally written by Fred Inbau and John Reid, and now in its fourth edition, provides a set of protocols known as the “Reid Technique.” The Reid Technique instructs police to suggest that the suspect has something to gain by confessing and that there will be consequences for not confessing. Police use storytelling and offer the suspect a series of alternative narratives. For example, they try to get the suspect to agree to having committed legally excusable or less reprehensible acts. They lie to suspects or use polygraph machines, not to detect lies, but to secure a confession.

Yet, the Reid manual is firm on the importance of never feeding crucial facts. If the suspect is told how the crime happened, police lose the chance to test the suspect’s knowledge of the crime. Because police have long known that suspects may admit to crimes they did not commit, for all sorts of reasons, including suggestibility, mental illness, desire for attention, and to protect loved ones, the Reid manual advises that “[w]hat should be sought particularly are facts that would only be known by the guilty person.” Leading questions are not to be asked, at least not as to crucial corroborated details. Instead, police ask open-ended questions such as “What happened next?” Further, police are trained not to leak the key facts to the media so that the public does not learn of them before trial.

Officers do not decide to interrogate a suspect lightly. Having decided to interrogate and try to secure a confession, they may contaminate a confession unintentionally or intentionally, out of a belief that the suspect is a danger to the public. If they do so unintentionally, they may not even realize that they have disclosed key facts to the suspect during interrogation. Traditionally, interrogations were not videotaped or recorded, yet interrogations can be long, complicated, and convoluted affairs. Washington, D.C., detective James Trainum has described how he unintentionally secured a false confession and did not initially realize what happened. There was a videotape of the entire interrogation, however, and when he carefully reviewed it, he discovered he had disclosed key facts to the suspect.

Confessions are incredibly powerful at a trial. As the U.S. Supreme Court has put it, “A confession is like no other evidence. Indeed, ‘the defendant’s own confession is probably the most probative and damaging evidence that can be admitted against him. . . . [C]onfessions have profound impact on the jury, so much so that we may justifiably doubt its ability to put them out of mind even if told to do so.” Arizona v. Fulminante, 499 U.S. 279, 296 (1991). Like Lloyd’s confession, those of the 40 individuals who were convicted yet later exonerated by DNA testing were especially powerful precisely because of their detail. All but two of the 40 told police more than “I did it.” Police reported that those people gave detailed and accurate information about the crime, including “inside information” that only the true culprit could have known.

Although defense lawyers have tried with varying skill and effort to challenge almost all of the confessions of those who were eventually exonerated by DNA evidence, judges found the confessions voluntary and admissible. The facts contained in confession statements were the central support for the state’s case. Just as in Lloyd’s case, judges repeatedly denied relief, relying on the seeming reliability of these confessions. Like Lloyd, after years passed, these defendants were exonerated by the failure of their DNA to match crime scene DNA. In most of those cases, confessions, certainly contaminated, may have been inadvertently contaminated. We cannot know precisely what happened in these cases, however, because they were not recorded. In 23 of the 40 cases (58 percent), interrogations were partially recorded. However, they were recorded only in the sense that Lloyd’s was “recorded.” The conclusion of the interrogation is recorded, not the process of getting to that conclusion.

Signs of False Confession

In most of these false confession cases, there were warning signs. When given a chance to volunteer information, the defendant got the facts wrong because he did not know anything about the crime aside from what he was told. In at least 75 percent of these cases (30 of 40 cases), the accused supplied facts during the interrogation that were inconsistent with the known facts in the case. Lawyers and experts analyzing confessions focus on such “lack of fit” when they assess whether a confession may have been contaminated. Had these interrogations been recorded, perhaps lawyers could have made stronger arguments that the confessions were, in fact, contaminated.

When defense attorneys did try to challenge those confessions, the trial judges rejected the challenges. After all, the confessions were accurate as to certain key details. The defendants supposedly knew facts that “only the killer could have known.” The Supreme Court has held that unreliability is not relevant to the question of whether a confession statement is sufficiently voluntary to be admitted at trial. However, judges often rely on the seeming reliability of the confession when finding the confessions admissible.

All of the 40 individuals who falsely confessed but were subsequently exonerated had waived their Miranda rights. Lloyd had been interrogated by the police while he was involuntarily committed in a mental hospital. Lloyd’s defense lawyer challenged the admissibility of the confession. The judge ruled that Lloyd’s confession was voluntary, noting, “[A]fter he is advised of his rights, it doesn’t make a difference what form it takes.” The Supreme Court’s voluntariness standard regulates interrogations more broadly by examining the “totality of the circumstances” to assess whether the confession was coerced, focusing on “the characteristics of the accused and the details of the interrogation.” Schneckloth v. Bustamonte, 412 U.S. 218, 226 (1973). Therefore, the judge in Lloyd’s case told the jury to consider “whatever circumstances surround[ed] the taking of the statement.”

Yet, the trial judge in the Lloyd case discounted the other indicia of potential involuntariness. One factor the Supreme Court has long taken into account when analyzing voluntariness is “the youth of the accused.” Schneckloth, 412 U.S. at 226. A third of those who confessed and were convicted and who were later exonerated for want of a DNA match were juveniles (13 of 40 cases; five were the defendants in the Central Park jogger case).

Another factor to be considered when assessing whether a confession is voluntary is the intelligence of the accused. Studies have long shown that individuals with low intelligence tend to be submissive and eager to please authority figures. They are peculiarly vulnerable to stress and pressure. Yet, at least 43 percent (17 of 40 cases) of those later exonerated and whose conviction was based on a confession were of low intelligence. Some, like Lloyd (who was involuntarily committed), were quite obviously of low intelligence.

Lloyd spoke to reporters about the case after his exoneration. The detective “provided me with quite a bit of information about the case,” Lloyd later recalled. Lloyd said that he had in fact heard about the green bottle in the neighborhood, but the other details came from the officers. Lloyd recalled, “He said, ‘What kind of jeans was she wearing?’ I said, ‘I don’t know.’ He said, ‘What kind do you think?’ I said, ‘Jordache.’ He said, ‘No, Gloria Vanderbilt.’” Lloyd said that the detective “guided him through a sketch of the garage” where the crime occurred, among other details. “The emphasis was on, ‘You want to help us, right?’ he said. I said, ‘Sure, I want to help any way I can.’” See Jodi Wilgoren, “Confession Had His Signature; DNA Did Not,” N.Y. Times, Aug. 26, 2002, at A1.

Jurors may have an understandably hard time believing that a person who is innocent could be coerced or manipulated into a false confession without some kind of torture. Therefore, it is advisable to call an expert to explain to the jury that false confessions do occur and that the defendant was psychologically vulnerable to police suggestion or coercion. However, judges often deny indigent defendants the funds to hire such experts, or they refuse to allow the introduction of such testimony.

After 17 years in prison, Lloyd was exonerated. He did not enjoy his freedom for long. He died two years after his release. Two civil rights actions were brought on his behalf, based on his wrongful conviction and inadequate counsel. Each was settled for not insubstantial sums ($225,000 and $4 million). As part of the settlement, the police entered a consent decree requiring that all interrogations in crimes that carry a penalty of life imprisonment without parole be videotaped. The wheels of justice grind slowly, but they do grind. Far more reliable convictions based on surer confessions will be ensured in Detroit. The police will have a record against which to test whether they are in fact suggesting facts to the suggestible.

In the 1990s, before DNA exonerations were so numerous, the only states to require the videotaping of confessions were Alaska and Minnesota. Many more states have followed that lead, but more needs to be done. Currently, 12 states and the District of Columbia either require or encourage electronic recording of at least some interrogations. Seven more state supreme courts require or encourage the recording of interrogations. At the local level, more than 750 police departments now videotape interrogations. Police departments have reported positive experiences with videotaping. They say that recording does not discourage a suspect’s cooperation. The videotape can simplify disputes over whether facts were fed or a confession coerced. As one federal judge put it, “I don’t know why I have to sit here and sort through the credibility of what was said in these interviews when there’s a perfect device available to resolve that and eliminate any discussion about it.” Transcript of Record at 72, United States v. Bland, No. 1: 02-CR-93 (N.D. Ind. Dec. 13, 2002) (hearing on motion to suppress).

The Detroit police chief commented, after the settlement in Lloyd’s civil case, “It’s a protection for the citizen that’s being interrogated. But from a chief’s point of view, I think the greatest benefit is to police because what it does is provide documentation that they didn’t coerce.” Peters, supra.

Recording entire interrogations alone is not enough. Just because an interrogation has been recorded does not mean that it was not coercive or unreliable. Indeed, the recording may show that there was coercion or contamination through the disclosure of facts. Judges should carefully review recordings to assess the voluntariness and reliability of interrogations. Police should adopt precautions for interrogations involving vulnerable juveniles or mentally disabled suspects. A new focus on accuracy can help to safeguard the reliability and legitimacy of confessions.

What do these false confessions tell us about wrongful convictions more generally? Contaminated confessions can produce trial evidence that seems very powerful to judge and jury alike. A confession is a seemingly surefire piece of evidence; it can convict the innocent. Videotaping all confessions will go a long way in reducing this danger. The rest is up to scrupulous police work and an increasing awareness on the part of judges and defense lawyers that even a seemingly voluntary confession can be contaminated. After all, cases like Lloyd’s show us how, without the benefit of DNA tests, the case against an innocent person may appear irrefutably strong.


Brandon L. Garrett

Brandon L. Garrett is a professor of law at the University of Virginia School of Law and the author of Convicting the Innocent: Where Criminal Prosecutions Go Wrong.