Glendale Case Raises Issue of Reliability of Confessions

For many people--including potential jurors--respiratory therapist Efren Saldivar’s alleged confession to ending the lives of as many as 50 terminally ill patients could be damning and persuasive evidence that he did it.

But such an assumption is not always sound, especially when, as in Saldivar’s case to date, there is no corroborating evidence, say experts who study the interaction between police and suspects.

Indeed, figures released by the state Wednesday indicate that the death rate at Glendale Adventist Medical Center actually declined during the bulk of Saldivar’s employment there, from 3.8% in 1989 to 2.1% in 1996.

As the investigation continued, it was revealed that Glendale police had subpoenaed records from Pacifica Hospital of the Valley in Sun Valley, where Saldivar worked part time for six months.

The lack of evidence so far suggests to some that Saldivar’s confession might have been fabricated. False confessions are surprisingly common, experts say, either from mentally unbalanced people who are looking for recognition or punishment, or from innocent people who are egregiously intimidated by the police.

Consider the 1991 slaying of six Buddhist monks, a nun and two acolytes in Phoenix.

Mike McGraw, 24, phoned police from a mental hospital, said he committed the crime and named three supposed accomplices. After two days of intense grilling, all three confessed and were on their way to the death chamber for the monstrous crime.

They were saved only when police accidentally arrested two youths possessing guns linked to the murders. The youths eventually confessed themselves.

The incident should be an object lesson to those who implicitly trust the veracity of confessions, said sociologist and lawyer Richard A. Leo of UC Irvine. With psychological bullying and other tricks of the trade, he said, “The police can make just about anybody confess to anything.”

“Even someone who is highly trained and intelligent can be susceptible to police intimidation,” added Dr. Howard Zonana of Yale University, medical director of the American Academy of Psychiatry and the Law. “It’s very hard to resist it under the heat and pressure of the situation.”

Known Cases Called ‘Tip of the Iceberg’

No one knows how many false confessions occur every year. But two studies several years ago--one that examined 70 proven cases of wrongful imprisonment in England and one that looked at 350 such cases in the United States--found that false confessions were one of the leading causes of overturned convictions.

And because many, if not most, such cases go unnoticed, Leo said, “the ones we do know about are just the tip of the iceberg.”

False confessions have been a particular problem in England, where, until recently, convictions could be obtained solely on the basis of confessions without any corroborating evidence.

The law was changed, in large part, due to the reversal of several notorious terrorism convictions, such as those of the Guildford Four (documented in the 1993 movie “In the Name of the Father”), who spent 15 years in custody for pub bombings, and the Birmingham Six, who were incarcerated 16 years for two 1974 bombings that killed 21 people. All the confessions were later shown to have been coerced by brutal police treatment.

Glendale police were unable to charge Saldivar because U.S. law requires that confessions be supported by corroborating evidence.

According to sociologist Saul Kassin of Williams College in Massachusetts, who along with Leo and psychologist Richard Ofshe of UC Berkeley is one of the leading U.S. experts in this field, there are three main types of false confessions:

* The voluntary confession, without any solicitation from police.

* The coerced confession, in which a suspect admits a crime under the pressure of interrogation but later recants.

* What Kassin calls the coerced internalized confession, in which an innocent person being interrogated comes to believe that he or she actually committed the crime.

The voluntary confession is the one most people are familiar with. McGraw’s confession in Phoenix is a good example. Others include the more than 200 people who confessed to kidnapping the Lindbergh baby and the large number who confessed to the 19th century London murders committed by Jack the Ripper.

Such confessors, experts say, often have low IQs or are mentally ill. Psychologist Ralph Pohlman of Toronto cited a patient who confessed to him that she had been responsible for the deaths of several babies in the cardiac unit at Sick Children’s Hospital in the mid-1980s, even though she had never even been in the building.

Many such innocent confessors feel they need to be punished for transgressions they did commit, perhaps during childhood. Others are seeking celebrity, which can come instantly in today’s mass media age. Still others may be covering for someone else.

Police usually protect themselves against such people by withholding some details of the crime from the public. But even that technique is not always successful, as evidenced by the Phoenix case.

The second, and more disturbing, category involves people who confess simply to end the interrogation. “The subject knows he is innocent, but goes along because he is led to believe that the short-term benefits of confessing outweigh the long-term costs,” Kassin said.

Thirty or 40 years ago, that interrogation might have involved billy clubs and brass knuckles. Today, police use sophisticated psychological techniques with aftereffects that may be less visible to defense lawyers, but nevertheless are as effective as physical methods.

“It’s a wild ride once you get into the interrogation room,” Kassin said. Leo added: “The more serious the crime, the rougher the interrogation.”

Most police departments rely on a manual called Criminal Interrogation and Confessions, which provides detailed instructions on how to win a confession from a suspect. The approach that the manual details is also called the Reid method.

Using this technique, police may start by being friendly, but eventually they will take a harder line, manufacturing witnesses and other evidence linking the suspect to the crime, such as DNA test results and fingerprints. They may hook the suspect up to a lie detector, or even a fake lie detector, and say the person is failing it.

“Police are free to invent evidence, because the courts believe that innocent people will not confess falsely,” Leo said. “Of course, that is what happens all the time.”

Police may also withhold something the suspect desires, such as food, medication, sleep or access to a phone. Kassin cites a teenager who was picked up for questioning after school and held for hours. Interrogators would not allow him to phone his presumably worried mother to tell her where he was until he confessed to the crime. After he did so, and called his mother, he immediately recanted.

The Reid technique even encourages the solicitation of false confessions. The suspect in a murder case, for example, might be told that interrogators think it was an accident, provoked or, perhaps, self-defense--even though they know it was not. If the person confesses to that, then they try to escalate it into a murder confession.

Finally, they will hint that the prisoner will be charged with a lesser offense or receive more lenient punishment if he or she confesses, and a harsher charge or penalty if the denials continue. They may even hint that the subject can go free if he confesses--which usually does not happen.

When such confessions are presented in court, it very often turns into a “he said, they said” case, in which police veracity is pitted against the suspect’s claims of coercion. Juries almost always believe the police, experts say.

When Innocent People Believe They’re Guilty

The third type of confession, and the most intriguing to psychologists, occurs when “an innocent person--anxious, tired, confused, subject to highly suggestive interrogation--actually comes to believe that he committed the crime,” Kassin said.

Such victims are often young, of low intelligence, highly trusting of the police, extremely suggestible and under great stress.

In 1978, 18-year-old Peter Reilly came back from a Methodist youth meeting to his house in Canaan, Conn., to find his mother slain. The likable, easygoing youth was clearly distressed when he called for an ambulance, but police brought him in for nine hours of questioning.

After gaining his trust, they eventually told him that he had failed a lie detector test. “Transcripts of the interrogation reveal a remarkable transformation, from adamant denials to confusion and self-doubt, then to complete conversion” and confession, Kassin said.

Reilly spent two years in prison before independent evidence proved he couldn’t have been the killer.

Even more startling is the case of Paul Ingram, a sheriff’s deputy in Olympia, Wash., whose daughters allegedly regained memories of childhood sexual abuse on a religious retreat in 1988. Eventually, the charges encompassed Satanic rituals in which he was said to have murdered infants in his backyard.

Ingram was interrogated 23 times over five months, hypnotized, told that he failed lie detector tests and given graphic details of the supposed crimes. His minister was brought in to urge him to confess. He was told that sexual offenders like him often repressed the details of their crimes.

Eventually, he confessed to the Satanic crimes, pleaded guilty and was sentenced to prison, where he now resides. “But there is absolutely no corroborating evidence to suggest that the crimes even occurred,” Kassin said. “There are no missing babies, nothing.” And, other experts say, “recovered” memories of childhood molestations have been almost completely discredited.

Kassin said he has often asked detectives how they can risk using such powerful techniques on innocent people. The response, he said, is always, “We don’t. We can tell, from body English and other signs, who is guilty.”

Kassin added that he is conducting tests to see whether investigators actually can discriminate between the innocent and guilty that way. “So far, we are finding that they can’t,” he said.

But Cmdr. Dave Kalish, spokesman for the Los Angeles Police Department, said that “false confessions are not that common” during police investigations. “Occasionally, during a high-profile crime, somebody might come out of the woodwork, but that doesn’t occur all that often.”

In response to criticisms, a growing number of police departments videotape or audiotape the entire interrogation so that defense lawyers and others can be aware of potential problems.

Police have not said whether their three- to four-hour interrogation of Saldivar was recorded.

Times staff writers T. Christian Miller and Matt Lait contributed to this story.