Prof. Paul Cassell

No one says that police have brought out the rubber hose, but some of their interrogation techniques are raising questions about why innocents confess to crimes they didn't commit.

Two days after 12-year- old Stephanie Crowe was found stabbed to death on the bedroom floor of her Escondido, Calif., home, her 14-year-old brother, Michael, told police he had killed his younger sister as she slept.

Not long afterward, one of his friends and classmates told police he stood lookout the night of Jan. 20, 1998, while Michael and a third boy sneaked into Stephanie's bedroom and killed her.

Based entirely on the two boys' statements, police came up with the theory that Michael, jealous of the attention his sister was getting, enlisted his friends in a plot to kill her.

The only thing missing in the investigation was the murder weapon. Police thought they found it when they came across the younger brother of one of the boys playing with a hunting knife. Later the older boy told police he had been given the knife after the killing and told to hide it.

With that, all three were arrested on charges of conspiracy and first-degree murder, and the case became an overnight media sensation: Three seemingly typical teenagers from apparently "good" homes whose shared fascination with knives and computer games had gotten way out of hand. One TV tabloid show even went so far as to describe the case as "The Devil Son Who Murdered the Angel Daughter."

The only problem with the whole scenario is that all three boys may be innocent. That became clear at the start of jury selection in a trial in January, when prosecutors disclosed that Stephanie's dna had been found on a bloody sweatshirt worn by a 29-year-old vagrant seen knocking on doors and peering in windows the night the girl was murdered.

In late February, prosecutors dismissed all charges against the three, though they left open the possibility that they may be refiled later. By late May, the vagrant had not been charged in connection with Stephanie's death.

If the boys hadn't joined in a plot to kill Stephanie, the "confessions" by two of them, who both quickly recanted, raise troubling questions about human nature. But they also raise troubling questions about standard police interrogation tactics.

What kind of person would confess to a crime he or she didn't commit? And given the fact that some people apparently do, what makes them more likely than anybody else to confess falsely?

The answer, experts say, is that while practically anybody can be made to confess to something he or she didn't do under the intense psychological pressures of a modern police interrogation, the young and the mentally impaired are more susceptible than most.

That's because the young and the mentally impaired tend to be more suggestible, more eager to please, more deferential toward people in positions of authority and less capable of rational decision-making than the average person.

The whole point of an interrogation, after all, is to extract a confession from somebody police suspect is probably guilty by leading that person to believe the evidence is overwhelming, his or her fate is certain, and the benefits from confessing outweigh the costs.

To do that, police interrogators employ a variety of psychological tactics designed to wear down a suspect and break the person's resistance--from appealing to the suspect's conscience to fabricating claims of evidence--all of which is perfectly legal and highly effective.

Hopeless Confusion

But the same tactics that work so well at getting the guilty to confess sometimes work just as well with the innocent, who tend to confess to a crime they didn't commit for one of two reasons. Either they come to the conclusion that their situation, while unjust, is hopeless and will only be improved by confessing, or their faith in their own memory is so badly shaken they come to believe they are guilty even though they don't remember it.

"The logic is really quite simple," says Richard J. Ofshe, a sociology professor at the University of California at Berkeley who studies the relationship between police interrogation tactics and false confessions. "If you put somebody in a situation he believes is hopeless and you give him a choice between two bad options, you can get him to say just about anything you want."

Ofshe recalls one case in which police induced a suspect to confess by telling him they had satellite photos of him committing the crime. In another case, he says, a quick-thinking detective invented a new type of bogus technology--called a neutron proton negligence intelligence test--to persuade a suspect that police had scientific proof he fired the gun used to kill two people.

Saul Kassin, a social psychologist at Williams College in Williamstown, Mass., who specializes in the dynamics of police interrogations, says average people tend to think they would never confess to a crime they didn't commit. But average people don't understand how stressful a police interrogation can be, he says.

"We all have our breaking point," Kassin notes. "When somebody reaches his or her breaking point, all he or she wants to do is escape. And the quickest means of escaping a police interrogation is to tell [interrogators] what they want to hear."

The Vulnerable Ones

The question, then, is not whether innocent people falsely confess, but how often.

E. Michael McCann, Milwaukee's district attorney, says any experienced prosecutor knows it can, and sometimes does, happen, particularly when the person confessing is a young child or someone with limited intellectual abilities.

But McCann says a false confession can usually be distinguished from a truthful one depending on whether it includes details about the crime only the offender would know. The danger, he says, is that the false confessor could inadvertently have been provided those details before or during interrogation.

"Even the most conscientious cop can make a mistake," McCann says. "That's why you have to be extremely careful in a situation like that, particularly if there is no independent, corroborating evidence to go along with the confession."

Some prosecutors admit there are false confessions and there are coerced confessions, but they say there is no such thing as a police-induced false confession.

"Innocent people do confess sometimes, which is a real problem for law enforcement," says Joshua Marquis, the Clatsop County, Ore., district attorney. "But the idea that somebody can be induced to falsely confess is ludicrous. It's the Twinkie defense of the 1990s. It's junk science at its worst."

Paul Cassell, a law professor at the University of Utah and a former federal prosecutor, studies the phenomenon of false confessions. He says the problem doesn't appear to be pandemic, as others have suggested, but confined to a very narrow and especially vulnerable subset of the population, namely the mentally impaired.

"The evidence suggests that those with mental limitations are at special risk of false confessions, but that it's actually quite rare and hardly ever results in a wrongful conviction," he says.

Yet Kassin says there have been enough documented instances of false confessions in capital cases, which tend to get far more scrutiny than noncapital cases, to suggest that the problem is a lot bigger than anyone would like to think.

Ofshe and a colleague, Richard A. Leo, a professor of criminology, law and society at the University of California at Irvine, claim to have identified more than 250 likely cases of false confessions in the post-Miranda era. Sixty of these have been documented in an article they wrote for the Winter 1998 issue of the Journal of Criminal Law and Criminology. 88 J. Crim. L. & Criminology 429.

In their article, Ofshe and Leo not only analyzed the evidence in the 60 cases they examined but grouped them into three categories: 34 of which they classify as proven false confessions, 18 as highly probable and eight as probable.

One case they classify as probable may even have resulted in an innocent man's exe- cution, the authors contend. That is the case of Barry Lee Fairchild, who was executed in 1995 for being an accessory in the 1983 abduction, rape and murder of a Pulaski County, Ark., woman.

According to Ofshe and Leo, no independent evidence linked Fairchild to the crime. And Fairchild, a mentally handicapped man who had steadfastly maintained his innocence, insisted he had confessed to the crime only because he had been beaten and tortured by the local sheriff and one of his deputies. (Two former sheriffs now admit that beatings were a common interrogation tactic at the time of Fairchild's arrest.)

A case Ofshe says is sure to appear on the next list surfaced in Chicago last August, when the police announced they had solved the July 28 murder of an 11-year-old girl, Ryan Harris, with the arrest of two boys, ages 7 and 8, both of whom were said to have confessed to the crime. But charges against the boys were dropped after semen found on the victim's underwear was linked to Floyd Durr, an ex-convict who was already awaiting trial for sexually assaulting three young girls. In late April, Durr was charged with Harris' murder as well.

Interrogation Tactics

But the Stephanie Crowe case is a textbook example of how a police interrogation can go wrong, according to Ofshe, who was an expert for the defense, and others connected with the case.

Police thought from the beginning that the murder was an inside job because the house showed no sign of forced entry. And they quickly settled on Michael as the prime suspect. One of the first officers on the scene had described him as being "inappropriately bereaved."

Michael, who was immediately separated from his parents, was brought in to the station for questioning the day after Stephanie's murder. Over the course of the next five hours he adamantly and repeatedly insisted he had nothing to do with his sister's death. But he was brought back the next day and administered a controversial lie detector test known as a computerized voice stress analyzer--which he was told he had failed--and then grilled for another six hours.

Police used every trick in the book to get Michael to confess, according to a transcript of the interrogation, most of which was videotaped.

They told him there was a mountain of evidence against him. They made him write a note of apology to his dead sister for having killed her. They told him his parents hated him and never wanted to see him again because they had come to believe he had killed his sister. And they told him if he confessed he would receive psychological treatment instead of prison, where, he was reminded, he would have to shower with some unsavory characters.

At one point in the interrogation Michael began to wonder aloud whether he might have had something to do with his sister's death, though he still maintained he had no memory of it. Eventually, he came to accept the notion that he must be guilty.

"I completely blocked myself out," he said near the end of the session. "And I wouldn't even know that I did it if she wasn't dead. It just as easily could have been a dream. I can't remember."

After Michael's confession, police repeated the same process with Joshua Treadway, the 14-year-old friend who supposedly stood lookout. After 11 hours of questioning, he not only confessed but implicated 15-year-old Aaron Houser.

"You'd have to know Josh to understand why," says Mary Ellen Attridge, Treadway's public defender. "He's very naive and very gullible. There's not a streetwise bone in his body."

Meanwhile, Richard Raymond Tuite, the 29-year-old vagrant and a diagnosed schizophrenic with a lengthy arrest record who is now serving a three-year sentence for an unrelated burglary, was questioned by police the day Stephanie's body was found.

But he was quickly discounted as a possible suspect. And while the red sweatshirt Tuite was wearing during questioning was confiscated, it wasn't until late last year, when Attridge spotted what appeared to be blood stains on one sleeve, that it was tested for dna.

The Crowe family has already filed a civil rights suit against the Escondido police and the San Diego County district attorney's office. And the Treadway and Houser families are said to be contemplating similar legal action.

"It's like one long nightmare," says Stephen Crowe, Michael and Stephanie's father, who bitterly accuses police and prosecutors of trying to put away three innocent boys to save their own behinds. "I don't know how they sleep at night."

Ofshe, who estimates that he has appeared as an expert witness in more than 125 false-confession cases, admits that his view of police interrogation tactics may be skewed by his experience with the practice. But he says the evidence he and others have amassed should persuade any fair-minded person that something needs to be done to reduce the likelihood of police-induced false confessions.

A requirement that all interrogations be recorded, which is now the rule in only two states--Alaska and Minnesota--would help, he and other experts say.

What the Justice System Can Do

But the real solution, they argue, would be to give courts authority to evaluate the reliability of a confession by comparing a suspect's account of the crime with the known facts. If the substance of the confession conformed closely with the facts, it would be admitted; if not, the confession would be suppressed.

But Cassell of the University of Utah isn't convinced that false confessions are as serious a problem as Ofshe and others suggest. Cassell doesn't believe giving the courts authority to scrutinize confessions is a good idea. The consequence, he suggests, would be suppression of more truthful confessions from the guilty.

Cassell says that his examination of the facts in nine of the 29 cases in which Ofshe and Leo claim a false confession may have led to the wrongful conviction of an innocent person shows that all nine confessors were probably guilty.

In the Fairchild rape and murder case, for example, Ofshe and Leo claim there was no independent evidence connecting the defendant to the crime. But Cassell says he found that Fairchild took police on a tour of the crime scene after confessing, and that a watch recovered from his sister, which Fairchild claimed to have bought in a pool hall, was identical to the victim's missing watch.

In eight of the nine cases he examined, Cassell says, the defendant appeared to be mentally retarded or suffering from serious mental problems.

The only solution, he says, is to videotape all interrogations and do away with the Miranda warning, which he says does nothing to protect the innocent and only makes it tougher to get truthful confessions from the guilty.

In the meantime, the innocent may have only their willpower to rely on. 

Mark Hansen is a reporter for the ABA Journal. His e-mail address is markhansen@staff.abanet.org.
Copyright American Bar Association. 

How the System Works
Truth in Justice