The Devil in Confessions.

It is a humbling experience to go to the Web site of the Innocence Project (http://www.innocenceproject.org), a nonprofit legal clinic at the Cardozo School of Law that handles cases in which postconviction DNA testing of evidence has provided proof of innocence. A recent visit to the site led me to the case of Eddie Joe Lloyd, who was wrongfully convicted of the murder of a 16-year-old girl in Detroit. During police interrogation, officers fed him information that he could not have known, such as details about the victim’s clothing and the location of her body. Lloyd confessed and was tried by a jury that convicted him after less than an hour’s deliberation. He was sentenced to life without the possibility of parole, and was freed in 2002 after serving 17 years in prison. He is the 110th American to be exonerated by DNA testing, and a good case to spotlight the false-confession problem. Lloyd’s is one of the 155 wrongful-conviction case profiles described on the Innocence Project Web site. Although the convictions in the vast majority of these cases appear to have been due to faulty eyewitness memory, about a fifth of the cases involved defendants who falsely confessed. Kassin and Gudjonsson have thoroughly reviewed the literature on confessions, providing scientific evidence bearing on just about any question you might ask about the subject, as well as identifying the very real problem of false confessions and recommending some solutions. They use case studies, archival data, results of laboratory and field experiments, and other forms of evidence to analyze confession evidence and its impact on real people and society. Their monograph is a superb example of psychological science in the public interest. People intuitively feel that they would never confess to something they did not do. But people do confess. They confess to things they actually did (in confessionals, in psychotherapy, and in police interrogations). And they confess to things that they did not do. One goal of our legal system must be to secure convictions of the guilty, but another must be to minimize wrongful convictions, including those involving false confessions. There is much about the legal process that traps the innocent in the confession net, and there are some ways we can, in principle, widen the holes of that nasty net. What I have always found particularly disturbing about the extraction of confessions by police is the use of a common interrogation tactic: presentation of false incriminating evidence. If the police had wanted to, they could have told Eddie Joe Lloyd that his fingerprints had been found at the scene or that an eyewitness saw him commit the murder. Such trickery and deceit is perfectly legal in the United States (although interestingly, in many European countries, lying to suspects is not permissible). To see why such a tactic is a problem, one has only to look at the false-memory literature and note what ordinary individuals can be led falsely to believe. In recent work, subjects have been deceived into believing (on the basis of a story experimenters said the subjects’ parents had provided) that, as children, they had been lost in a shopping mall for an extended time before being rescued by an elderly person and reunited with their parents. In other studies based on this lostin-the mall paradigm, subjects came to believe that they had had an accident at a family wedding, that they had been victims of a vicious animal attack, or that they had nearly drowned as children and had been rescued by a lifeguard. And in the famous computer-crash paradigm, developed by Kassin and his collaborators, subjects presented with false evidence that an eyewitness saw them hit a forbidden key on a computer keyboard were especially likely later to confess to having committed that prohibited act. So we have every reason to believe that some people who arc presented with false evidence that they committed a crime might actually come to believe that they did. In such cases of internalized false confession, people might not only confess to acts they did not do, but in some cases even confabulate false memories to go along with their confession, producing what is sometimes called a full confession—a detailed and convincing, but untrue, account of the crime and how it was committed. How often do the police actually use this type of trickery? One study of the interrogation tactics most frequently observed in 182 actual police interrogations suggests that it happens about 30% of the time. Studies using the computer-crash paradigm have taught us much about false confessions in an experimental setting. They have taught us some people will make false confessions and come to believe in their own guilt even when their confessions have substantial financial consequences. And they have taught us that teenagers will confess falsely at greater rates than adults. And from the studies of actual inmates, we learn some of the reasons why people confess. Although some suspects confess because they are psychologically manipulated into believing PSYCHOLOGICAL SCIENCE IN THE PUBLIC INTEREST