Imagine a small, windowless room. The fluorescent light hums overhead. A clock on the wall ticks away minutes that feel like hours. You have been sitting in a hard chair for a very long time, answering the same questions over and over again. The detectives across the table seem calm, patient, even friendly at times, but there is a weight in the air. You are tired. You are confused. You have asked for a lawyer, but the questions keep coming. At some point, a terrifying thought crosses your mind: maybe if I just say what they want to hear, they will let me go home. This is the kind of moment that lies at the heart of a new study from the University of Mississippi. Criminology professors Mary Catlin and Thomas Bryan Smith wanted to understand when and how police interrogation tactics can produce false confessions, and their findings, published in the Journal of Criminal Justice, add a crucial layer of nuance to a long and often painful debate. They found that the likelihood of a false confession grows steadily with the length of an interrogation. That may sound like common sense, but the numbers behind it are sobering. False confessions are not everyday events, but when they happen, they are catastrophic. They ruin the life of the person who confesses, and they leave the real perpetrator free to commit more harm. The stakes could not be higher, and this study helps explain exactly why time matters so much.
To understand the significance of this research, it helps to look at the broader picture of wrongful convictions. Recent studies suggest that wrongful convictions account for roughly 4 percent of all capital cases in the United States, and people who are wrongly convicted spend an average of eight years in prison before they are released or exonerated. That is eight years of lost freedom, lost family time, lost opportunities, and immense psychological damage. There are many reasons why innocent people are convicted, but the National Registry of Exonerations reports that 13 percent of all exonerations are tied to false confessions. That is a significant share of a problem that already shakes public confidence in the justice system. Earlier this year, another study estimated that police interrogation tactics were associated with false-confession wrongful convictions in about 1 percent of cases. At first glance, 1 percent may seem small, and in some ways it is. But in the criminal justice system, small percentages translate into real human beings. The researchers, however, had some concerns about that earlier analysis. Thomas Bryan Smith pointed out that the vast majority of interrogations are short, routine, and do not involve aggressive tactics. If you put every interrogation into a single bucket, the guilty confessions from those common short interrogations will far outnumber the rare false confessions that come from long, intense interrogations. The average becomes misleading. So they decided to ask a different question: what happens if you separate interrogations into different buckets based on how long they last?
Using data from previous studies, Catlin and Smith ran thousands of simulations to estimate the likelihood of a false confession leading to a wrongful conviction at different interrogation lengths. The results were striking. At a typical interrogation length of about an hour and a half, the risk of a false confession wrongful conviction is only about 1 percent. But at 12 hours, that risk jumps to as high as 25 percent. At 34 hours, the risk could reach 50 percent. Let those numbers sink in. A person who is questioned for a few hours is overwhelmingly likely to be telling the truth, even if they confess. But after more than a day of interrogation, the odds of a false confession become dangerously high. This is not just an abstract exercise. Studies show that interrogations that yield false confessions average about 16 hours, and some last longer than a full day. Imagine being awake for 16 hours, being isolated, being confronted with supposed evidence that you know is wrong but cannot explain, being told that things will go easier if you just cooperate. Your judgment starts to erode. Your sense of reality begins to blur. You may begin to doubt your own memory. In that state, even an innocent person can say things that are not true simply to make the ordeal stop. The researchers are quick to note that these are modeled probabilities, not observed rates, but the pattern is unmistakable. The longer an interrogation lasts, the greater the risk of a catastrophic mistake.
Duration, however, is only one piece of a much larger puzzle. Mary Catlin stressed that you have to account for stress, mental health, social persuasion techniques, and so many other factors. Interrogations are not neutral conversations. They are deliberately designed to break down resistance, and that breakdown can happen faster for some people than for others. A person with intellectual disabilities, a young person, someone with mental illness, or someone who is simply exhausted and terrified may be especially vulnerable. Many exonerees later say that they confessed because they believed it was the only way they would be allowed to go home. That is a heartbreaking statement. They did not confess because they committed a crime. They confessed because they wanted the pain to end. They wanted to see their families. They wanted to sleep. In that moment, a false confession can seem like a rational choice, even though it will destroy their lives. The impact is massive. An innocent person goes to prison, sometimes for decades. The real criminal remains free, perhaps to hurt others. The victim’s family is left with a false sense of justice. The police officers involved may genuinely believe they did the right thing. And when the truth eventually comes out, it is often too late. The damage is already done. This is why the debate over interrogation tactics is not an abstract academic argument. It is about protecting the most basic promise of the justice system: that we do not punish the innocent.
The researchers are also careful to warn against overinterpreting their findings. They are not claiming that their simulation is the final word, nor are they saying that the earlier estimate of 1 percent is wrong. Instead, they argue that both studies are based on modeled probabilities, not observed rates, and that the available data is simply not precise enough to support firm conclusions. Smith put it plainly: these simulation studies are too preliminary to be used to inform decisions in individual cases. A simulation can be informative, but it is still only a simulation. It should serve as a guidepost for what future research needs to do, not as a blueprint for current policy or as evidence in a courtroom. This is a remarkably humble position, especially in a field where people often want clear answers. But that humility is exactly what is needed. If policymakers or judges latch onto a single number and use it to justify a new rule or a legal decision, they risk making things worse. The danger is not in acknowledging that long interrogations can produce false confessions; the danger is in pretending we know exactly how often that happens when we do not. What we can say with confidence is that the risk exists, that it grows over time, and that it deserves serious attention. We need more research, better data, and a willingness to question practices that have been taken for granted for too long.
So where do we go from here? Catlin and Smith point to decades of research showing that evidence-based approaches to interrogation produce more accurate results than accusatorial tactics. There are proven methods that reduce the risk of false confession and, in some cases, actually increase the likelihood of obtaining a true confession. These methods focus on building rapport, asking open-ended questions, avoiding prolonged sleep deprivation, and steering clear of deceptive tactics that convince innocent people that they have no way out. Some jurisdictions have already adopted these approaches, and they have worked. Catlin asks a simple and powerful question: if we know these methods exist, and they have been adopted in some places, why would we not use them? There is no good answer. The goal is not to make it harder for police to do their jobs. The goal is to help them do their jobs better. A confession that is obtained through ethical, evidence-based methods is more reliable, more likely to hold up in court, and less likely to destroy an innocent life. The new study is a reminder that time is not neutral. Every hour in an interrogation room increases the risk of a devastating mistake. But we can choose to be smarter. We can choose to protect the innocent while still holding the guilty accountable. That is not a radical idea. It is just justice.