The dangerous psychology of how false confessions rewrite witnesses’ memories

Seeing is not always believing.
Seeing is not always believing.
Image: Reuters/Luke Macgregor
We may earn a commission from links on this page.

It’s one of the puzzling truths of wrongful conviction cases: Where was the evidence proving the wrongfully named suspect’s innocence? If he swears he wasn’t at the scene, surely there’s an alibi confirming where he was?

A growing body of evidence shows that when suspects make false confessions—which happen for many reasons, and far more often than most people realize—fiction trumps fact. Our brains abhor ambiguity. The tidy narrative of a confession biases witnesses and investigators alike, neutralizing evidence that could prove a suspect’s innocence.

In one study, 61% of people changed their identifications in a lineup once they knew a different suspect had confessed. Forensic scientists are more likely to erroneously identify a suspect if they know the person has confessed.

“It kind of creates this perfect storm of bias,” said Jeff Kukucka, assistant professor of psychology at Towson University. “Everyone involved is well-intentioned, but because they are human and therefore fallible, these biases create an unfavorable outcome for an innocent person.”

Kukucka is an author of a new study (pdf), published online in August and forthcoming in the journal Law and Human Behavior, which finds that a suspect’s confession, even if recanted, can even rewrite the memories of alibi witnesses.

A research team set up an experiment to trick subjects into believing they were part of a crime scene. Two students—one a study subject, the other secretly working for the researchers—were left alone in a room with some cash, before moving to an adjacent room to work in separate cubicles.

A researcher then approached the subject to say the cash had been stolen and the other participant was suspected. More than 90% of subjects confirmed the suspect was with them in the other room during the time of the alleged theft, audible in the next cubicle.

The subject was then told one of three things: that the suspect maintained their innocence, that the suspect confessed but recanted, or that the suspect recanted—and that continued insistence on the alibi made the subject suspect too.

When the suspect maintained his or her innocence, 95% of alibis stood by them. But when told simply that the suspect confessed and recanted, 55% backed off their earlier statement. And when subtly threatened, as often happens in police interrogations, a full 80% of alibi witnesses withdrew their support.

“When these people were changing their story, it’s not that they were becoming a little more reluctant. They were confidently changing their story,” Kukucka said. In some cases, “they were going from ‘I am sure she didn’t leave’ to ‘I am sure she left.’”

It’s hard to believe that anyone would confess to a crime she didn’t commit. But people confess all the time to things they haven’t done, because of coercion, confusion, exhaustion, or the belief that it will lessen their sentence for an inevitable conviction.

Since 1989, 31% of the 337 people in the US exonerated post-conviction because of DNA evidence made false confessions or incriminating statements, according to the Innocence Project. In homicide cases, 63% of exonerees had confessed.

There is one clear solution hailed by lawyers, psychologists, and law enforcement alike to prevent false confessions—electronically recording police interrogations, from start to finish. Given the corrosive effects of false confessions on justice, the best defense is preventing them in the first place.