I Did Not Do It, But I Said I Did: The Neuroscience of False Confessions

In the spring of 1989, a woman was brutally attacked while jogging in Central Park. Within days, five teenagers had been arrested. Within weeks, four of them had confessed on video. The confessions were detailed. They were emotional. They were played for the jury.

They were also entirely false.

The Central Park Five, as they came to be known, spent between six and thirteen years in prison before DNA evidence identified the actual perpetrator, a convicted rapist named Matias Reyes who had no connection to any of them. Reyes confessed in 2002. The five men were exonerated. [1]

What happened in those interrogation rooms is not a mystery. It is a documented psychological and neurological process, and it happens far more often than most people are willing to believe.

What a False Confession Is

A false confession is a statement in which a person admits to a crime they did not commit. This seems, on its face, like something that should almost never happen. The assumption most people carry is that an innocent person would never confess to a crime, because there is no rational reason to do so.

That assumption is wrong, and the science explains why.

Researchers who study false confessions have identified three distinct types. A voluntary false confession occurs when someone comes forward without any pressure, sometimes to protect another person or to seek notoriety. A compliant false confession occurs when a person who knows they are innocent confesses anyway because the immediate pressure of the interrogation becomes unbearable and confession seems like the only way out. An internalized false confession is the most disturbing: it occurs when a person comes to genuinely believe, as a result of interrogation, that they must have committed the crime, even when they have no memory of doing so. [2]

All three types exist. The compliant and internalized varieties are the ones most relevant to understanding what interrogation does to the brain.

Lens One: The Scientific Evidence

The Innocence Project, which has used DNA evidence to exonerate hundreds of wrongfully convicted people, has found that false confessions played a role in approximately 29 percent of wrongful convictions in its case database. That makes false confessions the second leading contributor to wrongful conviction in the United States, behind only eyewitness misidentification. [3]

Research by psychologists Saul Kassin and Gisli Gudjonsson, two of the leading scholars in this field, has documented the conditions that reliably produce false confessions. Long interrogations are among the most powerful. Studies of documented false confession cases have found that the average interrogation in those cases lasted more than sixteen hours. The Central Park Five were interrogated for between fourteen and thirty hours, mostly without a parent present, before they confessed. [4]

Other contributing factors include youth, which is particularly significant because adolescent brains are still developing the regulatory systems needed to resist psychological pressure. Sleep deprivation, which the structure of overnight interrogations often produces, also dramatically increases the likelihood of a false confession. And the use of deception by interrogators, including false claims about evidence, is a third factor that research has consistently linked to false admissions. [2]

Critically, all of these factors are legal. In the United States, police are permitted to lie to suspects about evidence. There is no time limit on interrogations. And for much of the twentieth century, there was no requirement to record them at all.

Lens Two: The Neuroscience

To understand why prolonged interrogation can produce false confessions, you need to understand what the interrogation environment does to the brain.

The standard interrogation technique used by law enforcement in the United States for most of the past century is called the Reid Technique. It places a suspect in a small, controlled room, presents evidence of guilt as a certainty even when it is not, offers psychological justifications for the crime to minimize the perceived consequences of confessing, and applies sustained social and psychological pressure until a statement is obtained. [5]

From a neuroscience perspective, this environment is designed, whether intentionally or not, to push the brain into a state where rational long-term decision-making becomes extremely difficult.

When a person perceives themselves as trapped and under threat, the amygdala activates the stress response. Cortisol and adrenaline rise. The body prepares to fight or flee. But in an interrogation room, neither option is available. The result is a state of sustained, inescapable stress that research has shown to severely degrade the functioning of the prefrontal cortex, the brain region responsible for long-term planning, consequence evaluation, and the ability to hold firm to a position under social pressure. [6]

Sleep deprivation compounds this effect dramatically. Even a single night of missed sleep produces measurable impairments in prefrontal cortex function that are comparable, in some studies, to being legally intoxicated. A person who has been awake for twenty-four hours while being subjected to sustained interrogation is making decisions with a brain that is functioning far below its baseline capacity. [7]

For adolescents, the situation is more severe. The prefrontal cortex is not fully developed until the mid-twenties, which means that teenagers have less regulatory capacity to begin with. When that already limited system is further compromised by stress and sleep deprivation, the ability to make rational long-term decisions, like the decision not to confess to a crime you did not commit, is severely diminished. The four teenagers in the Central Park Five who confessed were between fourteen and sixteen years old. [1]

The internalized false confession adds another layer of neuroscience. Research on memory and suggestion has shown that under conditions of high stress and sustained pressure, the brain can be led to construct false memories. Elizabeth Loftus, whose work on the misinformation effect was discussed in our post on eyewitness testimony, demonstrated that entirely fabricated memories can be implanted in willing research subjects with relatively modest suggestion. Under the much more intense conditions of a lengthy interrogation, the susceptibility is far greater. A person who begins by knowing they are innocent can, through this process, end by genuinely believing they are not. [8]

Lens Three: The Legal Interpretation

The legal standard for a valid confession is that it must be voluntary, meaning free from coercion. Courts evaluate voluntariness by looking at the totality of the circumstances surrounding the confession.

The problem is that the totality of the circumstances standard was developed at a time when the neuroscience of stress, sleep deprivation, and decision-making did not exist in its current form. Courts applying the standard have generally permitted confessions obtained after very long interrogations, after the use of deceptive evidence claims, and after sustained psychological pressure, as long as no overt physical force was used. [9]

The Supreme Court's decision in Miranda v. Arizona in 1966 established the now-familiar requirement that suspects be informed of their right to remain silent and their right to an attorney before custodial interrogation. [10] Miranda was intended to address the coercive nature of interrogation. But research has shown that the warnings are frequently misunderstood, particularly by juveniles and people with cognitive impairments, and that the psychological pressure of a Reid-style interrogation does not disappear simply because a warning was given.

Some reforms have taken hold. Many jurisdictions now require that interrogations be recorded in full, which provides a documentary record that courts can review. Several states have implemented special protections for juvenile suspects, including mandatory presence of a parent or guardian during questioning. The Supreme Court's ruling in J.D.B. v. North Carolina in 2011 established that age must be considered when determining whether a suspect was in custody for Miranda purposes. [11]

But the Reid Technique itself, and the legal permissibility of deception during interrogation, remain largely intact. The United States is one of the only developed democracies in the world that allows police to lie to suspects as a formal interrogation strategy. The United Kingdom, Canada, and most of Europe have moved toward interrogation models focused on gathering information rather than obtaining confessions, and research suggests these models produce more reliable results. [5]

Putting All Three Lenses Together

The science is clear that false confessions are not rare, irrational anomalies. They are predictable outcomes of interrogation conditions that push the brain beyond its capacity to make reliable long-term decisions. The neuroscience explains the mechanism: sustained stress degrades the prefrontal cortex, sleep deprivation compounds that degradation, adolescent brains are especially vulnerable, and the conditions of intensive interrogation can produce not just compliance but genuine false belief. The law permits nearly all of the conditions that produce these outcomes and evaluates the resulting confessions under a voluntariness standard that does not account for what is actually happening in the brain.

The Central Park Five eventually received a settlement from the City of New York. The detective who led the interrogations maintained for years that the confessions were genuine. The question of how a system that claims to seek truth can continue to use methods that neuroscience has shown to produce false ones is one that belongs at the center of any serious conversation about criminal justice.

A person who confesses is not always guilty. A confession that feels real is not always true. And a system that treats a statement made by a sleep-deprived, terrified teenager after thirty hours of interrogation as reliable evidence of guilt is not a system that has reckoned honestly with what we know about the human brain.

The next post will examine psychopathy, one of the most misunderstood diagnoses in forensic psychiatry, and ask what it actually means for criminal responsibility when someone's brain is wired in a way that fundamentally alters their capacity for empathy and moral reasoning.

Sources

[1] Burns, Ken, Sarah Burns, and David McMahon, "The Central Park Five," documentary, 2012; and New York v. Wise et al., exoneration records: innocenceproject.org

[2] Kassin, S.M. and Gudjonsson, G.H., "The Psychology of Confessions," Psychological Science in the Public Interest, 2004: journals.sagepub.com

[3] The Innocence Project, False Confessions and Recording of Interrogations: innocenceproject.org/false-confessions-recording-interrogations

[4] Drizin, S.A. and Leo, R.A., "The Problem of False Confessions in the Post-DNA World," North Carolina Law Review, 2004: scholarship.law.unc.edu

[5] Meissner, C.A. et al., "Interview and Interrogation Methods and Their Effects on True and False Confessions," Campbell Systematic Reviews, 2012: campbellcollaboration.org

[6] NIH, stress and prefrontal cortex function: pmc.ncbi.nlm.nih.gov/articles/PMC4316406

[7] Harrison, Y. and Horne, J.A., "The impact of sleep deprivation on decision making," Journal of Experimental Psychology, 2000: psycnet.apa.org

[8] Loftus, E.F., "Planting misinformation in the human mind," Learning and Memory, 2005: learnmem.cshlp.org

[9] Leo, R.A., "Inside the Interrogation Room," Journal of Criminal Law and Criminology, 1996: scholarlycommons.law.northwestern.edu

[10] Miranda v. Arizona, 384 U.S. 436 (1966): supreme.justia.com/cases/federal/us/384/436

[11] J.D.B. v. North Carolina, 564 U.S. 261 (2011): supreme.justia.com/cases/federal/us/564/261