Figuring out whether someone is guilty of a crime isn’t a straightforward task. Juries are often asked to reach a verdict in the face of unreliable eyewitness testimony and contradicting evidence. That ambiguity can lead to a shocking number of wrongful convictions, as dissections of high-profile trials in the NPR podcast Serial and the Netflix documentary Making a Murderer reveal.
But when someone confesses, a guilty verdict seems justified. No suspect would ever admit to a crime they didn’t commit … right? Guess again. Studies have shown that false confessions contribute to as much as a quarter of known wrongful convictions. Now, the latest work suggests that a good amount of those false confessions may be due to a common interrogation technique: sleep deprivation.
Interrogators sometimes resort to extreme, morally questionable measures to extract criminal confessions, including deafening noise, intense emotional manipulations and withholding food, water and rest.
“Many of these interrogations involve these extreme techniques,” says study coauthor Elizabeth Loftus, a psychology and social behavior professor at the University of California, Irvine. “Given that many people are often interrogated when they are sleepy after long periods of staying up, there is a worry that investigators may be getting bad information from innocent people.”
Around 17 percent of interrogations happen between the normal sleeping hours of midnight and 8:00 a.m. According to previous work, the majority of false confessions pop up after interrogations lasting longer than 12 hours, with many exceeding 24 hours. That suggests plenty of suspects are sleep deprived while they are being questioned.
In the new study, 88 participants were asked to complete a series of trivial computer tasks over the course of three sessions. At the beginning of each session, they were repeatedly warned not to press the “escape” key on the computer keyboard, or all the experimental data would be lost.
“To dissuade participants who may have been tempted to press the forbidden escape key, a member of the research staff watched as participants completed the computer tasks,” the authors write in their paper.
After the second session, half of the participants slept for eight hours while the other half were forced to stay up all night. The following day, all participants were told to sign a written statement in which they were falsely accused of pressing escape during the first visit to the lab. If they refused, they were given a second opportunity to confess to this fabricated crime.
The sleep-deprived subjects were 4.5 times more likely to falsely confess—50 percent of them caved in to the demands of the researchers, while only 18 percent of the well-rested subjects admitted to the wrongdoing, the researchers report this week in the Proceedings of the National Academy of Sciences.
When those strong-willed individuals who refused to sign were prodded a second time, the sleep-deprived subjects were 3.4 times more likely to own up to the crime—their numbers jumped to a total of 68.2 percent, while their rested counterparts rose to just 38.6 percent.
“There are a lot of cognitive tasks that are impaired when people are sleep deprived,” says Loftus. “Reaction times, judgment and problem solving, for example.”
Previous research also suggests that sleep deprivation impairs our ability to anticipate the consequences of our actions, to resist suggestive influences that might produce false and distorted memories and to inhibit impulsive behaviors. A subsequent analysis by the same team revealed that subjects who were naturally impulsive were more likely to falsely confess when sleep deprived.
For this study, the consequences were less severe than prison time—just the shame of potentially compromising the study-within-a-study. But Loftus believes the results still apply to crime fighting.
“We were interested in how the different variables affect the likelihood of confession,” says Loftus. “And I don’t have any reason to believe that sleep deprivation is going to affect behavior differently in this kind of a setting as compared to a real-world setting.”
So what motivates people facing more serious charges to confess to something they didn’t do?
“There are two types of false confessions that come about from police interrogation,” says Saul Kassin, a professor of psychology at Williams College who reviewed the study before publication. The first is a compliant false confession.
“These are situations in which people who know they are innocent reach their breaking point,” he says. “They are under stress and will do whatever it takes to escape the immediate short-term punishing situation—even if it involves a possible negative consequence later.”
The second is an internalized false confession, in which the innocent person not only confesses but actually starts to believe their own guilt.
“The police are allowed to lie to people,” says Loftus. “They tell them that their fingerprints were at the scene when they weren’t, that they flunked a polygraph when they didn’t, that an eye witness saw them do it when there is no such person. And these are powerful ways of getting people to believe what they are confessing to.”
Both these types of false confession are influenced by sleep deprivation, adds Kassin: “When people are mentally and physically fatigued, which is what happens in a sleep deprivation situation, they are more likely to do whatever it takes to end a punishing current situation than someone who has more mental energy to fight,” he says. “They are also more suggestible to misleading or false information about evidence that doesn’t really exist.”
People also sometimes falsely confess because they want the attention associated with a a high-profile crime. “That’s how you get 200 people confessing to kidnapping the Lindbergh baby,” says Loftus, referring to the infamous 1932 abduction and murder of aviator Charles Lindbergh‘s son. “But that’s obviously not going on in this experiment.”
Based on their findings, the researchers recommend that law enforcement officials evaluate suspects for their degree of sleepiness before an interrogation. The team also urged that all interrogations be videotaped so that judges, lawyers and juries can assess the value of the confession.
Still, law enforcement officials are unlikely to alter their tactics anytime soon, says Loftus: “There is obviously a belief that sleep-deprived interrogations help capture the guilty better. Otherwise this wouldn’t be used so frequently.”
Future work might investigate how sleep deprivation affects true versus false confessions, and how education, age and other demographics may influence the likelihood of a false confession from a sleepy suspect. The hope is that innocent people will get better protection, and investigators won’t waste any time finding the real criminals.
“Interrogation is a great process when everyone you interrogate is the criminal,” says Kassin. “The problem is, law enforcement doesn’t know in advance whether they are interrogating the perpetrator or an innocent person. They always think they are interrogating the perpetrator, but they may not be. And this is what makes it so important to protect against that worst-case scenario