Tell Me What I Want to Hear, Not What I Need to Hear: How Confirmation Bias Causes Wrongful Convictions
by Dale Chappell
What if I invited you to a conference at a fancy hotel where experts would present solid evidence that questions the validity of something you strongly believe, like maybe climate change? Would you come? What if those experts instead were to speak in support of your views, instead of challenging them? Now, would you come?
Most of us would go to the conference that supports our views. Why? Because we don’t like to be told we’re wrong. We seek confirmation in what we think and feel. It’s why we watch one TV news channel over another and why we read one newspaper rather than another. We seek out what we want to hear. Maybe it’s a survival instinct. But one thing’s for sure: We all suffer from this. It’s what experts call “confirmation bias.”
A recent study published in the Northeastern University Law Review by D. Kim Rossmo and Joycelyn M. Pollock, professors at Texas State University, exposed how confirmation bias infects criminal prosecutions. They analyzed 275 criminal cases overturned because of actual innocence and identified the most common causes of wrongful convictions in the top 50 of those cases.
What they found was that the thing we all deal with, confirmation bias, was the most common error that put innocent people in prison, sometimes for decades.
What Is Confirmation Bias?
In his book, How Judges Think, former U.S. Court of Appeals Judge Richard Posner defined confirmation bias as: “the well-documented tendency, once one has made up one’s mind, to search harder for evidence that confirms rather than contradicts one’s initial judgment.” Rossmo and Pollock further describe confirmation bias as a type of “selective thinking” and that once we form a hypothesis, our inclination is to confirm rather than refute it. They found that cognitive biases, such as confirmation bias, operate at a “below-conscious” level and are not intentional or deliberate decisions.
For example, the study concluded that confirmation bias could cause a detective to take evidence that supports his theory at face value without much thought, but then scrutinize evidence that contradicts his theory, looking for any reason to reject it. In other words, the detective is looking for information that is consistent with what he already believes.
This “top-down” processing is what Dr. Daniel Reisberg, chairman of the psychology department at Reed College, described as a person’s view being heavily influenced by their belief of what they are “supposed to see,” and that once this filter is in place, it’s hard to “un-see” that pattern and still be objective.
In short, you could say that confirmation bias is simply the desire to prove ourselves right, rejecting the fact that we might possibly be wrong.
Confirmation Bias’ Influence in Wrongful Convictions
The study found that confirmation bias was “the most connected causal factor by a significant margin” in wrongful convictions in criminal cases. They called confirmation bias the “pivotal position” in the cause of wrongful convictions. This is because faulty assumptions by law enforcement officers and prosecutors while trying to prove themselves right infected everything else that happened in the case, from what evidence got admitted to what witnesses were called or not called.
The study showed that confirmation bias, coupled with tunnel vision and a rush to judgment, switches a case from an evidence-based investigation into a suspect-based one, in an effort to solve the case. Soon there’s a “sunk cost” effect, they say, where detectives and prosecutors think they’ve invested too much into a case to even consider the possibility that they may be wrong. At this point, the study explained, even solid evidence that the suspect may not be the right guy gets discredited or outright ignored in order to not upset the momentum of closing the case.
How Does Confirmation Bias Influence Prosecutors in a Criminal Case?
Prosecutors are supposed to be an objective “check and balance” on who gets arrested and charged in a case, the study noted, but prosecutors suffer from confirmation bias the same as anyone else involved in solving a case. In fact, the study says prosecutors have even more of an interest sometimes in the outcome, and their confirmation bias therefore has an even more profound effect on the case.
Prosecutors are supposed to charge someone with a crime only if they have probable cause to believe they’re guilty. But because prosecutors often have a stronger connection to the victim of a crime, confirmation bias can lead to an “excessive zeal” to get a conviction, the authors say, despite evidence that might point in a different direction and to a different suspect.
Prosecutors are also trained to obtain convictions, the study pointed out. They count convictions as “racking up points,” making it difficult for prosecutors to ignore scoring points for their team (and their own career).
High-profile cases that cause a “media frenzy” can further drive prosecutors to do whatever it takes to resolve the case, the study showed. Even other prosecutors and law enforcement agencies can pressure a prosecutor to ignore crucial evidence and uphold a wrongful conviction. This almost happened in the 1989 “Central Park Jogger” case, when DNA evidence in 2002 exonerated the five men convicted of raping the jogger, after another man confessed and his DNA matched the crime scene DNA. Inexplicably, former prosecutors and New York Police Department detectives criticized the district attorney’s office for agreeing to drop the charges.
For prosecutors who hold their positions by election, an election year can be a driving force to aggressively pursue a conviction in a high-profile case. Take, for example, the Duke University Lacrosse Team rape case in 2007. Michael Nifong, the Durham County, North Carolina, prosecutor was in a hotly contested election, and the case presented him a chance to shine in the media.
Nifong made numerous public statements that the students were guilty, and he set out to prove this before all the evidence was gathered. Confirmation bias, the study’s authors said, caused Nifong to overstep bounds to secure a conviction, even ordering a lab technician to falsify a DNA report. Eventually, the State Attorney’s Office sent outside prosecutors in who dropped all the charges. Nifong was promptly disbarred.
In a disturbing case in 2004 in Wilmington, Illinois, the murder of a missing three-year-old girl was quickly pinned on the girl’s father by a prosecutor who was seeking re-election. The prosecutor promptly charged Kevin Fox, the girl’s father, with murder and announced that he was going to seek the death penalty. He also made public statements about the victim’s previous sexual abuse at the hands of Fox – which were false.
Fox sat in jail for eight months until a newly elected and different prosecutor dropped the charges, after a review of the case showed he didn’t commit the crime. Years later, the real murderer, Scott Eby, was found after his DNA turned up as a match in a DNA database. He was in prison by then for yet another murder and rape. Eby left other evidence at the scene, which investigators ignored because they were so focused on Fox: a pair of boots with “Eby” written on them.
Fox sued, and a jury awarded him $15.5 million after a seven-day civil trial, which was later reduced on appeal to $8.1 million in 2008.
In 1991, Westchester County, New York, District Attorney Jeanine Pirro, pursued the conviction of 15-year-old Jeffrey Deskovic, after police found one of his classmates raped and murdered in a park in Peekskill. The study’s authors called the investigation into Deskovic a “classic confirmation bias pattern”: when DNA evidence excluded Deskovic, prosecutors changed their theory of the case to continue focusing on him. They never considered that someone else committed the crime.
Instead, Pirro said Deskovic didn’t rape the victim but killed her in a rage after he caught her having sex with someone else. She said Deskovic was in love with the victim, even though the victim’s friends said he never had anything to do with her. Deskovic was convicted after he confessed to the crime during an interrogation, then sat in prison for 16 years before Barry Scheck, of the Innocence Project of New York, proved by DNA evidence that someone else had committed the crime: a man already in prison for rape and murder.
Pirro is no longer a prosecutor and is now a Fox News host who opines on criminal cases.
Preet Bharara was the U.S. Attorney for the Southern District of New York for over seven years before President Trump abruptly fired him in 2017. He said in his 2019 book, Doing Justice, that the cause of a wrongful conviction in a high-profile international terrorism case in Spain in 2004 was the result of a “culture conducive to confirmation bias” within the FBI. The arrest of Brandon Mayfield, a completely innocent person in Portland, Oregon, was nothing more than confirmation bias and a hesitation to challenge the senior agents’ first conclusions that Mayfield was their man. Despite that, officials in Spain told the FBI that Mayfield wasn’t the guy, they locked him up and insisted he was the terrorist they were looking for. Eventually, someone listened to reason, and they realized Mayfield had nothing to do with the terrorist attack in Spain. “An innocent man was accused and forever injured because of a failure to sufficiently reconsider,” Bharara said.
How to Prevent Confirmation Bias From Infecting a Criminal Case
The study found that 88 percent of the cases analyzed suffered from biased evidence evaluation. The authors suggested that better procedures for evidence collection, evaluation, and analysis would prevent many wrongful convictions. Of course, efforts to shield evaluation of evidence from confirmation bias was a major point for the authors.
The authors also suggested that training programs to “de-bias” investigators would help “mitigate” confirmation bias in criminal cases. Flawed decision-making because of confirmation bias was behind most wrongful convictions, the study found. While awareness of confirmation bias helps, training on better decision-making is what would help the most, they said.
The study further found that independent reviews of unsolved cases also would do much to cut down confirmation bias in criminal investigations. The authors noted a policy in the United Kingdom that after a certain amount of time, an unsolved murder is reviewed by a senior officer with no involvement with the case. If the case is high-profile, the officer is brought in from a different agency.
The study stressed that overall a wrongful conviction is an “organizational accident” with many small failures combined to produce the tragedy. No single “root cause” could be identified that if taken out would fix the problem. Confirmation bias, then, is a part of the problem, but not the whole problem, they said. But it’s a big part of the problem.
I have to admit, I noticed my confirmation bias in writing this story. Being an advocate for prisoners, I wanted to focus more on the errors of law enforcement and prosecutors in the numerous wrongful convictions identified in the study. I wanted to spotlight that they only care about one thing: winning at all costs, even if it means an innocent person would go to prison –or even the death chamber. Could you hear my bias? Even after consciously trying to be neutral, I could still hear it.
I learned that confirmation bias pervades all sides of every case. Taking steps toward reducing the effects of confirmation bias should be the goal of everyone involved in a criminal case: law enforcement, prosecutors, lawyers, judges, and even the media. Only then, I think, we’ll see a reduction in wrongful convictions and, as a result, more respect for the criminal justice system.
Additional sources: Rossmo, Kim and Pollock, Joycelyn, Confirmation Bias and Other Systemic Causes of Wrongful Convictions: A Sentinel Events Perspective (June 28, 2019). Northeastern University Law Review, Vol. 11, No. 2, 2019.