Expert: Many criminal confessions false
By Brent Curtis
Herald Staff | April 22,2005
When psychologist Saul Kassin first read the newspaper accounts of five young black men who confessed to a brutal rape in New York, he thought what most people would assume.
"I remember thinking the case was a slam dunk," said Kassin, a professor of psychology and chair of legal studies at Williams College.
The videotape he had of one of the young men's confession didn't do anything to change that assumption.
In a rambling interview, the man describes the grisly attack, including details about the scene, the act of dragging her off a trail and into a ravine, the knife used to cut away her pants and the large hand of one of his friends' clamped over her mouth.
He even finishes with an apology and the admonition "This was my first rape; it won't happen again."
That evidence was more than enough to convince a jury that Kharey Wise was guilty.
But Kassin told a gathering of students, law enforcement officials and others inside a Castleton State College lecture hall that Wise and his friends were innocent.
Still, the five men spent years in jail for their supposed roles in the Central Park jogger case — a highly publicized crime that involved a 28-year-old investment banker who was raped, beaten and left for dead in 1989.
Their innocence only became apparent in 2002 when the actual rapist confessed — an admission upheld by forensic, DNA and other evidence lacking in the false conviction.
That revelation generated numerous news stories, but what fascinated Kassin was the subject of his lecture Thursday: Why do innocent people sometimes confess to crimes?
Oftentimes, the answer lies in the techniques used by interrogators, he said.
Since the 1970s, the majority of law enforcement agencies in the country have relied on interrogation techniques developed by John Reid & Associates who developed plans for conducting interrogations and interviews.
Interrogations and interviews might sound similar, but the tone of the proceedings and the intent of investigators is vastly different, Kassin said.
Interviews are used to gather information and to get a "feel" for the people involved while interrogations are used to elicit a confession from a suspect who is presumed to be the guilty party.
"I used to ask 'How do you know you're not getting false confessions,'" Kassin said. "The answer always struck me as facetious, but it makes sense when you think about it: 'Because we don't interrogate innocent people.'"
But Kassin contended that investigators sometimes focus on the wrong clues.
Police officers are trained to pick out the truth tellers from the liars by observing a range of nervous behavior such as fidgeting, slouching, avoiding eye contact and other indicators, he said. But in tests conducted on college students and trained professionals, he said the average for picking out a liar was equivalent to a coin flip.
"Over time, investigators come to expect deception," he said. "They sometimes identify an innocent person as a liar based on mistakes they make."
Once in the interrogation room, Kassin said, there is little an innocent person can say to convince investigators that they aren't guilty.
Behind closed doors in a preferred 10-foot-by-10-foot space, investigators use a number of techniques to draw confessions from suspects, Kassin said.
But the interrogative tools that he said lead to the majority of false confessions are the legal misrepresentation of evidence by investigators, the implied understanding that a confession will allow a suspect to go home and, perhaps most of all, time.
While the elapsed time of an average interrogation is about one to four hours, most of the false confessions he's studied came after an average of 16 hours of questioning.
Kassin said the lack of time limits for interrogations was troubling, as is the practice — ruled legal by the Supreme Court in 1969 — of allowing investigators to "misrepresent" facts during interrogations.
"They can say 'We know you did it because the victim grasped the assailant's hair during the struggle and we have those hairs and the lab results show they're yours," he said. "I think that is deadly wrong. People can be induced to do any number of harmful things to themselves in that situation."
In fact, Kassin said misrepresenting evidence occasionally leads innocent people to believe they "must have done it" because police tell them they have key pieces of evidence pointing toward them.
But in most cases involving false confessions, the people being questioned admit to crimes they didn't commit just so they can get out of the interrogation room.
"They know they're innocent, but they just want to cut their losses and get out of a bad situation," Kassin said.
To prevent future false confessions, Kassin said he was a proponent for videotaping the entire interrogation process so prosecutors and judges can see how a confession was reached. He also said after the meeting that he hoped to refine the interrogation process in the future.
"My goal is, within the next 10 years, to help the police build a better mousetrap. One that is more surgically precise," he said.
As for the police, members of the Rutland Police Department's Bureau of Criminal Investigation gave Kassin's remarks mixed reviews.
"None of us want false confessions from people. We came here today to hopefully sharpen our techniques," Police Detective Ray Lamoria said. "But I would have liked someone from Reid to have been here for a point and counterpoint."
Police Detective Sgt. Kevin Stevens said he was troubled that Kassin didn't talk much about the physical evidence that police amass before interrogations.
"He didn't mention the physical evidence at all really," he said. "Before we interrogate anybody, we spend all of our time gathering physical clues that link suspects to crimes."
Detective Chris Kiefer-Cioffi said it was too easy for Kassin to be a critic of the interrogation process which she — and the other officers present — said has accounted for far more true confessions than false ones.
"I'm sure if he had the time to study the fallacies of the process then he sure must have some solutions, right?" she said.
Contact Brent Curtis at firstname.lastname@example.org.