On the night of 19 April 1989, a twenty-eight-year-old investment banker was raped and beaten nearly to death in Central Park. Within two days, five teenagers — Antron McCray, Kevin Richardson, Yusef Salaam, Raymond Santana and Korey Wise, aged between fourteen and sixteen — had given videotaped statements implicating themselves and each other.
The videotapes were the case. There was no physical evidence linking any of them. The semen recovered did not match any of them, a fact known before trial. The statements contradicted each other on nearly every material point: where the attack happened, who did what, what the victim was wearing, the sequence of events, even the time. Each boy placed himself as a peripheral participant and the others at the center.
They were convicted in 1990. In 2002 a convicted serial rapist named Matias Reyes, serving life for other crimes, confessed. His DNA matched the evidence. He had acted alone. The convictions were vacated in December 2002.
The interrogations that produced the tapes ran between fourteen and thirty hours. Almost none of it was recorded — the videotape began only after the statements had been shaped, so what a jury saw was the end product with the process invisible. Parents were present intermittently or not at all, and where present were themselves pressured. The boys were told, individually, that the others had already implicated them and that cooperating was the only way to go home as a witness rather than a defendant.
That last technique is the structural core of the case, and it is a prisoner's dilemma constructed deliberately. Each suspect is told the others have defected. Under that belief, confessing while minimizing your own role dominates continued denial — you get to go home, and the blame lands on people who have already blamed you. The choice is rational given the false information supplied, and the information is what makes it work.
Adolescence multiplies everything. Laurence Steinberg's developmental research shows that the capacity for weighing future consequences against immediate rewards is still maturing well into the twenties, and that the gap is largest under emotional arousal and in the presence of authority. A sixteen-year-old in hour twenty of an interrogation is making a decision with a system that is not built for it.
The Innocence Project's data on DNA exonerations records false confessions or incriminating statements in roughly twenty-nine percent of cases, and about half of proven false confessors were twenty-one or younger.
And then the corruption effect. Once the tapes existed, everything else bent around them. The DNA exclusion was explained rather than credited — the prosecution theorized an unidentified sixth attacker. Juries convicted despite it, because a confession is the most powerful evidence there is, and jurors do not discount it much even when told the conditions that produced it.
Which is the general finding. A false confession is not one bad piece of evidence in a case. It reorganizes the case around itself.
Cross-ref: Brendan Dassey chapter