On Friday, Brendan Dassey, the 26-year-old convicted murderer who became famous thanks to Netflix's documentary series Making a Murderer, learned that his conviction had been overturned, meaning he would be set free in 90 days unless prosecutors decide to retry his case.
Dassey was convicted along with his uncle, Steven Avery, of the 2005 murder of 25-year-old photographer Teresa Halbach. But the Netflix series cast doubt on the narrative of the case advanced by authorities, and last week, a judge ruled that Dasey's 2006 confession was unlawfully coerced.
The full four-hour-long interview with Dassey was uploaded to YouTube last year. It's agonizing to watch. The detectives repeatedly tell Dassey—then only 16 years old—to "be honest" and reassure him that they're on his side; the teenager seems lost. They ask Dassey to give them details, and he struggles, answering with questions, as if checking in to see if he's giving police the "correct" answer.
They try to get Dassey to say that Avery shot the victim in the head, asking, "What else did you do? Come on… something with the head…"
"He cut off her hair," Dassey offers.
In his ruling, US Magistrate William Duffin cited the interrogators' "repeated false promises [that he shouldn't worry], when considered in conjunction with all relevant factors, most especially Dassey's age, intellectual deficits, and the absence of a supportive adult, rendered Dassey's confession involuntary under the Fifth and Fourteenth Amendments."
But though those "relevant factors" make the conduct of the police in the case especially upsetting, the techniques used in Dassey's interrogation, like lying to him and assuring him he won't get into trouble if he confesses, are common, accepted, and legal practices in law enforcement.
"Police interrogation in America is fundamentally dishonest," says Richard Leo, a professor of law and psychology at the University of San Francisco and a leading researcher on police interrogation practices. "Police are allowed to lie all the time. They call lying about evidence 'ruses.' It's kind of Orwellian in the sense that it's not really a ruse. If I'm telling you I've got your fingerprints and your DNA and I have a surveillance video… and you're a kid and you don't know that police can lie about evidence, you're thinking, How would I be in a video?"
Combined with other manipulations, experts say, such tactics can drive an innocent person to confess to murder.
"Every one of us has a breaking point," says Steven Drizin, a law professor at Northwestern University and one of Dassey's appellate attorneys. "When the tactics make the suspect feel that continuing to assert innocence is futile, the suspect reaches a point of hopelessness and becomes easier to manipulate."
When the only way out is a confession, many take it.
Juveniles are especially vulnerable, says Drizin, because "they tend to make impulsive decisions that focus on short-term rewards rather than long-term consequences… A teenager might confess simply because he thinks that by confessing he can go home."
Most, if not all interrogations in this country, are some variation on the Reid Technique, which was developed in the 1940s by a man named John E. Reid as a more humane alternative to the intense and often physically violent interrogations of the previous decades, the so-called third degree.
The Reid Technique relies on a presumption of guilt based on either evidence or, as Leo says, "a gut hunch."
"You isolate somebody," is how the professor describes it, "and then you accuse them. Cut off their denials, confront them with real or made-up evidence, then present minimizing scenarios—suggest that if you confess, it's either not a crime, or it's not that serious."
The Reid Technique also emphasizes the importance of reading body language.
"Nonverbal behavior is more reliable than spoken words," according to a training article posted on the website of John E. Reid and Associates. "Facial expressions, of course, reflect the person's attitude… Eyes are windows to the soul."
But there is no evidence that's true, say critics of the method.
"They're completely wrong," says Leo. "Scientific research contradicts almost everything they say about detecting deception from cues. So they are simply wrong in how they advise interrogators to divine truth-telling or deception from body language and physical behavior."
Dassey was 16 when he was charged in 2006, and according to court papers, he was a "slow learner" with a lower than average IQ and enrolled in some special education classes in high school.
The two detectives who questioned him, Calumet County sheriff's investigator Mark Wiegert and Wisconsin Department of Justice special agent Tom Fassbender, weren't getting the answers they were looking for and pressed him further.
"We know," Fassbender said to Dassey. "We just need you to tell us."
Dassey had nothing to say.
Finally, Wiegart gave up and just said it, ""All right, I'm just gonna come out and ask you. Who shot her in the head?"
"He did," Dassey answered, referring to Avery.
When asked why he didn't say that before, Dassey responded, "'Cause I couldn't think of it."
The process, called "contamination" or "scripting," says Leo, is "the feeding of non-public details to the suspect" and getting them to parrot them back to authorities.
"The bad news is it's not against the law," says Leo. "Courts tend to look the other way."
One of the most significant protections for suspects is videotaping interrogations. Wisconsin, where Dassey was arrested, mandates recording.
"We're finally starting to win the videotaping battle," says Drizin. "When I started this work, only two states required the police to record interrogations. Now we're close to 25. Recording is critical because it is the only way to see whether the suspect provided the details of the crime that only the true perpetrator should know or whether the police fed them to the suspect."
Many other countries, like England and Germany, don't allow police to lie to suspects, according to Leo, who wonders if it might be time to reevaluate our practices here.
"It's a really big question: Do we need thoroughly dishonest police to get confessions and solve crimes?" he asks. "If we accept that, one consequence is that we will have a higher false confession rate than we need. Is it fair? Is it ethical?"
He suggests the Reid Technique, while progressive in the 1940s, could be regressive today, and that the future of questioning suspects could be a more investigative-style approach, an actual interview.
"Not guilt-presumptive," he says. "Not rush to judgment, not trying to muzzle somebody to the point where they don't talk and then to repeat your theory… But instead to encourage them to give information—and with finesse and less psychological coercion, try to get accurate information as opposed to confessions. It might be time."