Untrue Confessions

Orange County Register/October 10, 2004
By Edward Humes

Startling research strikes at the core of criminal prosecution. Eye witnesses, even confessions, turn out to be unreliable.

The detective surveyed the grim scene that had preempted his morning coffee - a 12-year-old girl's slim form sprawled in her bedroom doorway, brown hair still pulled into a neat pony tail, her jeans and purple T-shirt smeared with blood. He was steeling himself for the next step in the investigation: confronting his prime suspect.

Stephanie Crowe had been stabbed nine times as she lay in bed, her family sleeping down the hall as she died. Her mom awoke at one point to the sound of a door creaking, but the house cats were always making little noises on their nightly prowls, and so she rolled over and returned to her dreams and a lifetime of if onlys.

There were no signs of forced entry at the Crowe house. Inside job, the detective concluded.

There had been only one other killing in Escondido that year, but it was nothing like this horror. This was the sort of thing that ruined a good town, the sort of thing that had to be solved, and solved fast.

The detective didn't have to look far for a solution: The family had been in hysterics since finding Stephanie early that morning - everyone, that is, but her brother, 14-year-old Michael Crowe. He sat, expressionless, the detective would later recall, calmly playing a handheld video game. "Inappropriate grieving," the police later called it. His dark hair, parted in the middle, the detective noted, looked a lot like the strands he had spotted in Stephanie's cold fingers.

The detective and his partners at the Escondido Police Department knew what had to be done - hard, unpleasant but essential. By the time they completed their many hours of interrogation of Michael Crowe, along with two friends thought to be accomplices, the police had used lies, false promises, isolation from parents and attorneys, even threats of adult prison and predatory older inmates to persuade the teenager to drop his protestations of innocence .

The detectives told a sobbing, gasping, pleading Michael that they had blood and lie detector evidence proving beyond any doubt that he was the killer -- lies, but convincing to the 14-year-old. Then they suggested he might have a split personality, a good Michael who would never hurt a fly, and a bad Michael who had locked away the memory of poor Stephanie's murder, enabling him to sit there and believe himself innocent. If he would only just admit it, detectives promised, they could help him.

In the end, the police got what they needed to bring the boys to justice: confessions from Michael and a friend, and enough incriminating statements from the third boy to file murder charges against them all. Sure, the interrogations were rough, the videotapes painful to watch, but such was the difficult path to truth. And so another dreadful case of violent, antisocial teens in the era of Columbine was laid to rest.

Except, there was one hitch: The insistent voice of a slim, intense psychology professor from the University of California at Irvine kept saying, No, you've got the wrong guys.

Richard Leo, who has spent most of his career studying the dynamics of police interrogations - both the good and the bad - viewed more than 40 hours of videotape of Crowe, his friends and the Escondido Police.When he was through withhis analysis, he declared the police work in the Crowe case a textbook example of how not to question suspects, finding that it amounted to a form of "psychological torture" so coercive that the boys would have said almost anything to make it stop.

If you wanted a lab experiment designed to prove how to bully suspects into falsely confessing to crimes, Leo concluded in a case that began with Stephanie's death in 1998 and continues to this day, you couldn't do any better than what took place in that Escondido police interview room.

The boys were innocent, Leo asserted, confessions be damned. And he was right.

Confessions, DNA, fingerprints, eyewitness testimony, the word of crime victims: These are the gold standards of criminal investigation, the best and most convincing tools for bringing the guilty to justice. Who could doubt a person's guilt with such compelling evidence in hand?

The answer, for most of the last century, was: No one. But such seemingly ironclad evidence has come under a new scrutiny from a cadre of researchers at UCI's School of Social Ecology - psychologists, social scientists and attorneys, all working in that gray zone where the law, science and study of the mind intersect. Their research has broken new ground and established UCI as one of the premier schools for law and psychology.

Their work has changed the way DNA evidence is regarded, uncovering grievous mistakes by crime labs nationwide, compelling others to improve their practices, freeing innocents from prison and death row.

They have challenged the science behind a hundred years of fingerprint comparisons, rattling forensics experts worldwide.

Their research has altered the way we view eyewitness reliability and raised new questions about "recovered" memories of long-past crimes.

And a "jury lab" regularly reveals how well - or how poorly -evidence is understood by ordinary citizens -- and why normal people sometimes hear and see things that justdidn't happen.

As lawyers and expert witnesses, their cases have included O.J. Simpson's, Oliver North's, the Oklahoma City bombing, and the Catholic Church's child abuse cases. Along the way, says one of the nation's leading experts in law and psychology, John Monahan of the University of Virginia Law School, "UCI has leaped to the national forefront."

There are many reasons that UCI's law, crime and psychology program is now considered among the two or three best in the nation, Monahan and other experts in the field say. But most agree that the school's rise to prominence began in earnest in 2001, when psychologist and memory researcher Elizabeth Loftus came to town, streaming acclaim and controversy in her wake.

"Hawaii?" You want me in Hawaii?"

The nation's most sought-after courtroom expert on eyewitness memory silently ponders a quick trip to the 50th state, the suggestion rendering her momentarily speechless -- which is not the usual state of affairs for Beth Loftus. But this offer is unexpectedly alluring, dangled before her by an eager defense attorney calling from Honolulu: Could she please take a quick summer trip to paradise to testify in a murder case?

She gets at least one such call a day, a deluge that has put her on the stand more than 250 times in her career. She's just back from a homicide case against a cop in Phoenix, and she'll soon head to Boston, where she'll analyze the "recovered" memories of people who have, after many years, remembered being abused as children by Catholic priests. There's no time for more trips, even to Hawaii.

"No," Loftus finally tells her crestfallen caller, stifling a sigh. "I just can't do it."

Loftus is in demand for a simple reason: The 59-year-old psychologist is the ultimate memory detective.

For the last quarter century, she has hunted the sources and causes of false memories and revealed just how easily our remembrance can be manipulated. In Loftus' memory lab, people recall events that did not occur, recognize strangers as familiar faces, and recount in sensory detail experiences that they have never had. In an experiment for a recent PBS documentary, Loftus and her team persuaded actor Alan Alda that eating a hardboiled egg may have made him sick during a picnic when he was a child - something he had previously insisted never happened.

All it takes is a suggestive or misleading question from an incompetent or devious interrogator, and an entire memory can change without a person even knowing it, Loftus has found. The implications of her research have been profound, for when false memories involve more than picnic lunches and party encounters, when they influence crime witnesses, victims and suspects, then liberty, life and the integrity of the justice system are at stake. In ways no one ever expected but which are now the basis of countless research projects, Loftus's work has permanently altered the way our recollections are treated in court.

Elected this year to the prestigious American Academy of Arts and Sciences, Loftus is one of the 25 most cited psychologists in scientific literature. To say her arrival was a coup for the university, says Leo, who helped recruit her and whose own research is based in part on her discoveries, is an understatement.

"Beth Loftus is not just a pioneer. She is the pioneer in the field of eyewitness testimony and memory," says Steven Clarke, chair of the law and psychology program at University of California-Riverside. "Beth Loftus was the first to develop the scientific methodology to study the memory of eyewitnesses. . She's a magnet for talent. She has catapulted UCI to the top in law and psychology."

Over the course of her career, Loftus's work shattered the once widely held notion that the brain stores memory like some sort of organic computer, an analogy she has disproved in a series of clever experiments that showed just how easy it is to influence, contaminate and manipulate whatpeople remember.

She began experimenting early on with witnesses' memories of car accidents - showing them recreations or videos of collisions, then questioning them. She found that even varying simple descriptive terms in questions - asking, for example, how fast the cars were traveling when they smashed one another versus when they hit one another - had profound consequences.

People in the "smashed" group tended to remember higher speeds than the "hit" group, and to claim more often that they saw glass breaking during the accident, although there had been no broken glass.

She had proved that leading questions, even seemingly innocuous ones, could contaminate and alter memories of complex events. The results were even more dramatic when she experimented with offering outright misleading information to witnesses, such as suggesting there was a yield sign at the scene of a collision when there was actually a stop sign. Those who received the misleading information more often remembered the signs incorrectly than those who were questioned in a neutral way.

Blame for an accident could be shifted, Loftus discovered, just by asking questions in a certain way. And the resulting false memories were not necessarily uncertain - many people recalled them with conviction, even emotion, just as if they were genuine experiences.

In a series of follow-up experiments, she demonstrated what would later be called the "Lost in the Mall Effect," finding that, on average, as many as one in four people can be convinced through suggestive questioning that they experienced all sorts of traumatic events that never occurred - getting lost in the mall as a child and becoming terrified; witnessing demonic possessions; getting bitten by an animal.

The test subjects often went on to add their own rich, detailed accounts of these non-existent events that they now adamantly believed to be true: fictional rescuers, fictional medical treatment, fictional joyous returns home.

Even visceral, primal memories are not sacrosanct in Loftus' Memory Lab, where she experiments on implanting false food memories as the ultimate dieting tool. "Just think of what we could do for people who need to diet if they could be persuaded to remember that they dislike fattening foods." she says.

Her work, along with that of other like-minded researchers, proved critical to debunking the mass child abuse cases that sprang up around the country in the 1980s, including the McMartin preschool case and dozens like it from Bakersfield to Boston. Loftus, Leo and others have found that children are especially vulnerable to having their memories altered by leading questions.

Although prosecutors have sometimes criticized and grilled her, and some experts have tried to cast doubt on her lab work by asserting that it exaggerates memory problems in the "real world," the science behind Loftus' conclusions about eyewitness memory is widely accepted. She says she never felt particularly controversial -- until she began about ten years ago to take on what she derisively calls "the recovered memory crowd."

"That's when all hell broke loose," she says.

When a series of spectacular and emotionally charged cases began cropping up in the'90s based on the theory of "repressed" and "recovered" memories of past traumas and abuse, Loftus discerned a pattern: The long-lost memories were almost always revealed during sessions with therapists who were already inclined to believe in repressed memories as a cause of depression, bulimia and a host of other disorders. Not only did she find the scientific basis for this dubious, Loftus also came to believe that many "recovered" memories could be attributed to therapists who inadvertently used suggestive techniques with their patients.

"Except people were being sent to prison this time," she says.

Her work helped free George Franklin of San Mateo, perhaps the most widely publicized recovered memory defendant, who was convicted of murder after his adult daughter recalled duringtherapy that she had witnessed him kill her best friend when she was 6 years old, more than two decades earlier. Franklin served five years in prison before successfully appealing.

Loftus' testimony on behalf of Napa winery executive Gary Ramona, accused of abusing his daughter many years after the fact, helped him successfully sue his daughter's Irvine therapist, then persuade a jury that the "recovered" memories that destroyed his career and marriage had been created in the therapist's office.

Most recently, Loftus has gotten involved in the defense of the Archdiocese of Boston and Father Paul Shanley in cases arising from recovered memories of sexual abuse. The District Attorney in Boston recently dropped charges involving two of the alleged victims, whose memories seemed most subject to Loftus-style challenge.

But psychologists remain somewhat divided on the validity of recovered memories, and Loftus has aroused the ire of true-believers, particular those who consider themselves victims of long-past abuse - which his why she has been swatted with a newspaper by a fellow air passenger, and why a man spoke at a conference on repressed memory not long ago about longing to slash Loftus' tires.

Indeed, UCI owes Loftus' presence at the Irvine campus to a controversial repressed memory case.

While at the University of Washington, Loftus raised doubts about a therapist's claims that he had videotaped a woman in the act of recovering a memory of child abuse. The patient on the video, which was shown by her therapist at conferences nationwide, complained that Loftus had violated her privacy by attempting to investigate the claims.

Loftus had been a faculty member for 29 years at Washington, yet she recalls with undisguised bitterness that, following this complaint, university officials arrived at her office and seized her files with fifteen minutes notice, ordered her not to speak about the case, and began an investigation that lasted 19 months. She finally was exonerated by the university, she says, but racked up $30,000 in legal fees in the process, then was sued by the patient when she and another psychologist published an account of their work on the case (without ever publicly identifying the patient). The lawsuit is pending.

Leo heard about the controversy and suggested UCI's School of Social Ecology make Loftus an offer. His own research on false confessions has repeatedly turned up impressionable people who come to "remember" perpetrating crimes they didn't commit. He told her he felt UCI supported professors whose findings challenged conventional thinking - something he had experienced when his work came under fire for freeing people from prison.

Loftus accepted UCI's offer of a titled professorship and the chance to work for what she calls "the best law and psychology department in the country."

Now Loftus sees a "critical mass" forming at UCI, and though she hesitates to suggest her arrival was the catalyst for this, the evidence of her influence is hard to miss.

Just down the hall, Assistant Professor Jodi Quas attacks the question of witness reliability from another direction: Her work is geared towards finding ways to make witness memory more reliable, particularly in young children, by scrupulously avoiding the "Lost in the Mall" questioning techniques. Quas is working with juvenile justice authorities in Los Angeles and elsewhere to help young children testify truthfully and to help cops and social workers question them in a neutral way.

Two other young researchers drawn to the UCI law and psychology group this fall also are shaking conventional views on kids and the legal system. Elizabeth Cauffman of the University of Pittsburgh and Jennifer Skeem of the University of Nevada-Las Vegas are studying the reliability of a checklist many juvenile courts use to identify young sociopaths.

Judges and juvenile social workers have been desperate for a tool to help determine which kids can benefit from treatment and which should just be locked away , and the checklist has seemed to fit the bill. But Skeem and Cauffman are finding that it could easily mistake normal teen problems for adult-style sociopathy. If the checklist isn't working as intended, Cauffman and Skeem say, then kids who can be saved are, literally, being thrown away.

The growing reputation of the law and psychology faculty at UCI has led to an increase in the number of students applying to the university's School of Social Ecology, as well as an influx of grants to keep the research going. "We're like kids at the playground," says Loftus. "This is the place to be."

Richard Leo's work seems to tie together the varied threads of the work of the psychology and law researchers at UCI. His views on false confessions and their causes combine theories about altered memories, suggestive questioning, the effects of stress and trauma on kids, and the way the legal system treats the young and the mentally unstable.

Leo and his mentor, UC Berkeley Professor Richard Ofshe, have done some of the world's most oft-cited research on the false-confession phenomenon. Along the way, Leo has consulted on more than 500 criminal cases and testified in more than 100, primarily for the defense. (He also has done training for the Miami Police Department and other law-enforcement organizations.)

Interest in false confessions came accidentally for the 40-year-old Leo: While an undergrad at Berkeley in 1984, a student in his class was grilled by police for 16 hours in a murder case. The young man ended up confessing to killing his girlfriend.

No physical evidence pointed to Bradley Paige as the killer of Berkeley student Bibi Lee, who was found dead on a hiking trail, her skull badly fractured. The police zeroed in on him out of habit: Statistically, most murdered women are killed by their husbands or boyfriends.

The account of the crime police extracted from him did not match the evidence at the murder scene -- although it did match preliminary information the interrogators had in hand about the girl's death, information that proved to be wrong. In later years, Leo would show how this sort of flaw can be a prime indicator of a false confession, evidence that the police, not the suspect, provided the story line.

But at the time, confessions were rarely attacked in court, and such fine points seemed like nitpicking. He said he did it, the prosecutor argued, so what else did you need to know? Page was convicted of manslaughter and sentenced to six years in prison - a conviction that has stood despite compelling evidence uncovered later implicating a convicted serial killer seen in the vicinity of Bibi Lee's murder

The case, and particularly the tactics of the police and the discrepancies between the confession and the crime, left a deep impression on Leo. Such cases have been the subject of his research ever since.

Leo does not design and perform experiments as Loftus does, but instead uses real-world criminal cases to study confessions and interrogation techniques, making detailed comparisons between accounts of crimes given by "confessed" criminals and the objective facts uncovered by forensic investigations and eyewitness interviews. Confessions are such powerful evidence that police, prosecutors and juries will often overlook irrefutable evidence of innocence- somebody else's DNA or fingerprints at the crime scene - rather than disregard the confession, Leo has found.

One particularly tough case for him was in Stanislau County last fall, when Joseph Alan McCarty was convicted of manslaughter for killing his best friend's mother. Leo was called to the stand in the case to testify about techniques police use to extract confessions, and how they can lead to false admissions, particularly with young and impressionable people under stress. McCarty was 20 at the time. But Leo was barred from saying outright that McCarty's confession was, in his opinion, false.

The jury foreman would later say he and hisfellow jurors, while convicting McCarty, found Leo "completely credible," but that they ended up believing the confession was legitimate in this case. The jury never knew what Leo and everyone else in the courtroom knew: McCarty had passed a polygraph test in which he denied responsibility for the murder. That test, like Leo's opinion, was legally inadmissible.

"It is depressing at times," says Leo. "False confessions have led to more wrongful convictions than any other single type of evidence."

Leo helped reach a happier outcome two years ago in Wenatchee, Washington, where a modern-day witch-hunt for child molesters led to the arrest of 43 residents accused of abusing 60 children. Nearly 30,000 criminal charges were filed before the case began to unravel, as allegations of police misconduct, threats and coercion of suspects and child witnesses - including the lead investigator's own foster daughter - began to surface.

But by that point, more than a dozen indigent and developmentally disabled defendants had been persuaded to confess and plead guilty.

Leo strives to avoid emotional involvement in the cases he studies, but he was infuriated by the Wenatchee case, and he felt compelled to get involved. Leo's work proved pivotal in freeing several of the accused (as did Loftus', who also took on several Wenatchee cases). In one case, Leo was able to show how Doris Green, a mother of four, had been coerced and threatened into confessing. She was freed after serving four years of a 23 year prison sentence.

This spring, Leo and a coauthor published a study that uncovered 125 false confessions that occurred in five years, in which DNA or other conclusive evidence subsequently proved the confessed criminals innocent. These were not cases in which innocence was possible or likely, but absolutely certain - confessed "criminals" who had done nothing more than walk into an interrogation room.

Among Leo's most provocative findings:

• False confessions are most common inmurder cases, where the pressure is greatest to make an arrest. The high stakes of such cases do not make the police more careful about avoiding false confessions - just the opposite.
• A third of false confessions come from juveniles, "the most vulnerable" to police pressure, says Leo.
• The average interrogation in the125 false confession cases lasted more than 16 hours, compared to a typical police interrogation, which averages less than two hours.
• Almost 60 percent of the false confession cases studied were dropped by police or prosecutors before trial. But of those false confession cases that went to trial, 81 percent ended with a conviction -- with nine receiving death sentences.
• Most of the false confessions would not have been revealed without the advent of modern DNA testing - innocents would remain in prison, or face execution.

The Stephanie Crowe murder investigation in Escondido illustrated all of these factors in a single case: juveniles accused of murder, high pressure on the police to solve a sensational crime, marathon interrogation sessions, and a case that entered jury selection before it was abruptly dropped because DNA evidence pointed to an entirely different killer.

Leo says it was his most memorable and disturbing case by far. He interrupted his year-long book-writing sabbatical so he could testify in the case this past spring.

Part of Leo's work required him to view the 40-plus hours of interrogation tapes in the case. They are painful to watch - no parent would willingly allow their child to be subjected to the crushing pressure exerted on the three boys by the police. But the parents had no choice. They were kept away, with the Crowes threatened with arrest and with losing custody of their youngest daughter even as they mourned Stephanie's death.

In the videos shot inside the Escondido Policeinterrogation room, 14-year-old Michael writhes, screams, sobs, appears to nearly choke, and begs the police to please, pleasestop. He bangs his head against the wall and cries out, "Oh, God, Oh God, no." He curls into an upright fetal position and says over and over that he could not possibly have killed his sister, that he would remember it if he had.

But the detectives are relentless.

They lie to him, they claim a voice stress analyzer reveals him as a killer, they say blood and other evidence links him and his friends irrefutably to the crime. They imply prison and rape by older inmates will be inevitable unless he confesses and gets help. And, finally, Crowe cracks.

"As one watches Michael Crowe's deep and anguished cries as he is told, repeatedly, and comes to believe, that he killed his sister without any knowledge or memory of doing so, one sees the picture of psychological torture," Leo later wrote in a report to Crowe's lawyer. " In my professional opinion, the interrogations of Michael Crowe were psychologically brutal, coercive and highly improper."

As so often happens with impressionable and young suspects, Leo says, Crowe reached a point where he began to say whatever the detectives wanted to hear -- anything to make the interrogation end. He begins to express doubts about his own mind and memory, the detectives having convinced him that there is a mountain of evidence against him. At last he says on the tape that he may have done it without really remembering it.

The detectives declared this a triumph and called Crowe's words an admission of guilt. Leo calls it a "coerced-persuaded" confession - when the confessor doubts his own memory and makes an admission based on the "facts" the police give him. The only real evidence in the case against Crowe was this confession, a similar one from one friend and incriminating statements from another.

The detectives failure to give Miranda warnings eliminated some of this taped evidence, and the rest of the case evaporated as trial was about to begin. It turned out that the Escondido Police had in their possession for months a shirtseizedfrom a 34-year-old schizophrenic transient, Richard Tuite, who witnesses saw in the neighborhood on the afternoon and evening of the murder. Police discounted him as a suspect as they focused solely on Stephanie's brother and his two friends.

At the insistence of defense attorneys, an independent lab finally examined Tuite's shirt. DNA tests identified Stephanie's blood spattered on the cloth. The police had the key to the case in their possession all along.

As for the hairs on Stephanie's hands that seemed so suspicious to detectives: They were stray hairs the girl probably picked it up from the carpet while trying to crawl out of her room.

In the wake of the DNA evidence, the San Diego District Attorney dropped the case against Crowe and his friends but refused to charge Tuite, saying prosecutors still believed a case could be re-filed against the boys. Leo says the power of the confession evidence swayed the authorities as they began to search for some theory that could put the boys and Tuite together -- even though none of the confessions mentioned a transient.

The Escondido Police have never publicly apologized nor retreated from their belief in the boys' guilt, and the city has succeeded in winning dismissal of most, though not all, of a federal lawsuit filed by the boys' family.

Police officials routinely decline comment on the case, but several of the original detectives served as consultants for a true-crime book that suggested the boys were more likely Stephanie's killers than Tuite. But a new investigation by the San Diego Sheriff's Department placed blame squarely on Tuite.

The state Attorney General took over the case, filing murder charges and calling on Leo as an expert witness to explain why the boys' confessions could not be believed. "It is one of the most egregious examples of inept and improper questioning I have ever seen," Leo testified.

Escondido detectives testified for the defense and in support of the confessions.

In May,six years after Stephanie died and long after the three boys on the confession videotape had become young men, jurors rejected the confessions and the detectives who extracted them, and voted to convict Tuite of voluntary manslaughter - a verdict both Leo and the Crowe family view as a form of vindication. Tuite received a maximum sentence of 13 years in prison in August, and faces additional time for briefly escaping custody during jury selection.

The attorney general's office was so impressed with Leo that they want him back to do training for their office about false confessions. "He's normally a defense witness, of course," says Deputy Attorney General Jim Dutton, a prosecutor on the Tuite case. "But he's the kind of witness who says the same thing no matter what side calls him Our job is to get it right, not just win a case. If there's a false confession, we want to know it."

If there's a common goal for the researchers in psychology and the law at UCI, they say it's not what some in law-enforcement tend to imagine: that they're out o prove America is overrun by rogue cops and bullying prosecutors. Rather, their findings suggest that injustices tend to arise from the best, not the worst, of intentions, from genuine - if misguided - desire to protect the public, from conventional wisdom that is anything but wise, and from good-faith beliefs that are, nevertheless, as false as the memories they can generate.

The power and continuing impact of their work is not simply a mater of pointing out how the authorities can get it wrong, say Leo, Loftus, Quas and the others. It's in helping the authorities get it right.