Search About Newsletters Donate
Adam Sirios, left, raises his hand to confirm he was the lone holdout juror, after a mistrial was declared in the murder case against Pedro Hernandez at the Supreme Court in Manhattan on May 8.
Justice Lab

Untrue Confessions

Experts say that people admit to crimes they didn’t commit. Why did only one juror in the Etan Patz murder case believe them?

In part because he trusted social science, Adam Sirois is now the most famous juror in America.

He was the lone holdout for a “not guilty” verdict in the Etan Patz murder trial, which ended with a hung jury. The case revolved around whether or not to accept defendant Pedro Hernandez’s on-camera confession to police that, in 1979, he lured six-year-old Etan to the basement of a SoHo bodega and strangled him. Hernandez was arrested in 2012 after the police received a tip that he had admitted to killing a boy.

In court, the Hernandez defense team, hoping to raise doubt about these confessions, presented evidence that Hernandez had been diagnosed with schizotypal personality disorder and an extremely low IQ. Those issues came up repeatedly during jury deliberations. “False confessions do exist,” juror Sirois, who works in public health, told the New York Post in a post-trial interview. Sirois mentioned expert witnesses in the case who described research showing that mentally ill and low-IQ individuals, like Hernandez, are more likely to admit to crimes they did not commit. “I really felt that’s what happened with Mr. Hernandez,” Sirois concluded.

Although the other 11 jurors were not ultimately swayed by the social science evidence on false confessions presented by the defense team, this body of research is, in fact, robust. One defense witness was Icelandic psychologist Gísli Hannes Guðjónsson, whose “suggestibility scale” tests whether an individual is prone to describing false memories. The test is administered by reading the subject a paragraph containing 40 facts. The subject is asked to state what he remembers about the paragraph. He is then asked a series of questions about the facts he heard, including some leading questions that suggest wrong answers. Low-IQ individuals, like Hernandez, earn higher “suggestibility” scores. They are more likely to recall incorrect facts and to change their recollections in response to leading questions.

The defense also called psychologist Bruce Frumkin, an expert on false confessions who conducted a series of tests on Hernandez in 2013. Frumkin testified that the defendant frequently “remembered” falsehoods, saying the boy he killed had dark skin (Patz was white) and incorrectly identifying his own age when he met his wife. Frumkin’s research has found that low-IQ individuals are prone to trust police and to misunderstand their Miranda rights, including the right to remain silent.

Why was only Sirois convinced by this research? Another set of studies shows that juries seem to question the core findings of false confession experts. According to social psychologists who have assembled mock juries to consider hypothetical cases, even those jurors who recognize that police questioning can be highly coercive — and who are theoretically aware that false confessions happen — are unable to shed the gut instinct that false confessions don’t make sense.

“Most jurors think no one in his right mind would do that,” said Richard Leo, an expert on false confessions at the University of San Francisco School of Law. “Most jurors are positive they wouldn’t do that themselves.”

As a result, according to research by psychologist Saul Kassin of Williams College, confessions produce a higher conviction rate among jurors than any other type of eyewitness or character evidence. Even when jurors are informed that a confession was coerced, they are unable to discount it when deciding on a verdict.

Leo says that overcoming a confession is particularly difficult for jurors in the context of American culture, which highly values individuality. “In our understanding of things, the person most authoritative about a person is the person himself,” he said. “We think of confession as going to church to tell about yourself, whereas confession in the police context is often about passively repeating what they have fed you.”

Perhaps jurors are more likely to credit social scientific findings that confirm common sense, as opposed to findings that contradict it. The concept of confessing to a murder one did not commit is profoundly difficult for many people to imagine. In contrast, the idea that adolescents are less responsible than adults is widespread. In 2005 and 2010, the Supreme Court cited emerging research on adolescent brain development in its decisions to roll back the juvenile death penalty and juvenile life without parole. Ever since, neuroscience has become a bigger part of defense strategies to mitigate guilt. In Boston, for example, Dzhokhar Tsarnaev’s lawyers called on Jay Giedd, a psychiatrist and brain researcher, to argue that a 19-year old’s ability to resist peer pressure and thrill-seeking is not fully developed. That evidence was a key part of the defense strategy to blame the older Tsarnaev brother, Tamerlan, for leading Dzhokhar astray, and spare the younger brother from the death penalty. The jury is still out — literally — on whether that argument was persuasive.

Defense attorneys and the experts they hire to testify are not yet sure how to get juries (and some judges, for that matter) to pay attention to more complex and counterintuitive social scientific findings, like those on false confessions. But Leo says that it begins with basic communication.

“It’s hard for us experts to communicate in a context — the jury trial — in which we can’t control the quality of the lawyers, the judge, the jurors,” he said. “But when we have our chance, we have to speak in a down-to-earth, non-academic way, so that we relate to these jurors and our important theories get through to them.”