Adam Benforado, a professor of law at Drexel University, describes these unsettling problems with the justice system in the recently published book “Unfair: The New Science of Criminal Injustice.” The book uses psychology and neuroscience to examine and expose the illogical and unfair ways that judges, jurors, attorneys and others in the legal system make decisions about who is sent to prison, and who walks free.
Benforado’s research shows that mistakes in the criminal justice system are more common than we like to think, and that our personal biases play a disturbingly strong role. He also argues that there are clear and easy steps that we could follow to limit these injustices, if we care to take them. This interview has been edited for length and clarity.
Your research looks at applying cognitive psychology to law. Why is that important?
Our criminal justice system has been built up over centuries and really millennia. Unfortunately when we look at the scientific evidence, we find that a lot of the underlying assumptions are not backed up by science. A lot of our legal system is based on incorrect assumptions about human behavior.
Can you give us an example of a case where assumptions made in the courtroom led to a verdict that was clearly wrong?
The one that stopped me in my tracks was a case involving John Jerome White. This was a brutal rape case from 1979 in Meriwether County, Georgia. We have the image from the line-up that was conducted in this case. The victim was brought in. She looked at these five men and she picked out the one in the middle. John Jerome White said he didn’t rape the woman, but he was sentenced and ended up spending a couple decades in prison.
Finally, the DNA from the rape case was tested – we didn’t have the capability back in 1979, but a couple decades later we did. And John Jerome White was not the attacker.
What really startled me about this particular image was the actual perpetrator was in that line-up. He was actually locked up for another offense, and they pulled him in at the last minute. What this shows is that this victim looked eye to eye with the man who attacked her and picked out the man standing next to him.
That shows that good people who have every incentive to get things right can make terrible, terrible errors which are extremely costly. They are costly not just to the person who is wrongly convicted, but to the victim -- not only have you suffered this horrible crime, but you are now responsible for putting an innocent person in prison. And, in this particular case, that actual perpetrator ended up going off and raping another woman in the intervening decades.
For me, that really summed up what this entire book was all about – the fact that the dangers to our community are not evil, greedy, malicious people, they’re our friends and neighbors. Those are the people who are ultimately responsible for some of the terrible injustice that is happening.
In 1979, we didn’t have DNA testing. Wasn’t this line-up the best option? Was there a better solution?
In that case, there were a number of things that were problematic. The woman had already looked at a set of images of potential suspects a few weeks earlier. One question is, was she remembering the person who raped her, or was she just remembering this image that she had seen? We know from laboratory research that just seeing someone’s image on Facebook can cause a person to be more likely to pick that suspect out of a line-up.
It’s also quite possible that some of the things the police were doing were highly suggestive. The woman’s initial description was of a man who was round faced with a stocky build. John Jerome White is rail thin. The only man who fits the description of the perpetrator is the actual perpetrator.
The lesson is clearly that we should rely on eye witness identification a lot less than we do. We know that tens of thousands of people are charged with crimes after being identified by eye witnesses. And we also know that a third of the time, eye witnesses in real identification pick out one of those innocent fillers. That’s a terribly high error rate, and that suggests we have to find other ways to identify perpetrators.
Research suggests that both the race of the victim and of the defendant influence sentencing. One study by researchers at Cornell found that defendants with more stereotypically “black features” were more likely to be sentenced to death. What do you think of that research?
That was a really powerful study. We’ve known that race was a problem with the death penalty for a long time. Far more people on death row are African American than would be predicted. And studies suggest that the factors that are meant to predict whether someone receives the death penalty don’t predict that – people who commit the worst crimes are not the ones who receive the death penalty.
What was so fascinating about this study is it showed that it’s not just if you’re black or white, it’s how black you are. So people with thicker lips, wider noses and darker skin were more likely than other African-Americans with less stereotypically African-American features to receive the death penalty. In my mind, even if you would otherwise support the death penalty, this is just proof that we are not able to administer it in a way that is fair and just under the rules of the constitution.
And it’s not just race, right -- thinness and attractiveness affect our judgments, too?
All interpersonal differences, which are not meant to have any impact on any aspect of trial, end up shaping how things go for defendants. With respect to witnesses, we know that attractive witnesses are more likely to be believed. We also know that weight has an impact. In one study, women who were overweight or obese were treated much more harshly by mock jurors than people who were thin.
A very important point is that a lot of these biases are operating beyond people’s awareness or control. So when it comes to implicit racial bias, for example, it’s not that judges who end up giving African Americans higher bail hate black people or are secret bigots. It appears that they are just susceptible and have been exposed to the same negative stereotypes linking blackness and violence or crime that we all have been exposed to.
Those damaging stereotypes end up having an impact on real-world behavior. There’s interesting research that doctors offer different medical interventions based on the color of people’s skin. We know that recruiters look at a resume differently if it has an African-American name or a white name. In all these cases, the people that are engaging in this discriminatory behavior, it appears, are not doing so out of racial animus. They are doing it because they are engaging in automatic behavior which has been engrained over a lifetime of being exposed to a culture in which African American lives are devalued and blackness is coded with a lot of negative imagery.
What are other measures that we could take to improve decisions? I know that you mention “virtual trials” – what are those?
One of the things that I suggest in the book is that we need to control biasing factors. In essence, if we know that jurors and justices can be swayed by the attractiveness of witnesses, or we know that jurors place a lot of weight on whether the defendant is making eye contact, or his hands are shaking on the witness stand – if we know from research that that is not a good way to tell whether someone is lying, it makes sense to begin to control for those factors.
The defendant’s skin color shouldn’t make a difference in the outcome. The prosecutor’s mannerisms or bombastic style shouldn’t make a difference. And yet from research we know that all these things are biasing factors. So why not control for these things by eliminating judge’s and jurors abilities to see the color of the defendant’s skin, or the mannerisms of the prosecutor?
The technology to do this actually already exists. We conduct business transactions in virtual space. We can do heart surgery without being in the same room as the patient. So why not think about the potential of virtual adjudication for the future?
In terms of benefits, it’s not just that this would prevent judges and jurors from biased determinations. It would also change the behavior of attorneys. One of the big problems we have in the way we conduct trials is we allow attorneys to strike jurors. In some parts of the country, black jurors are commonly kept off death penalty juries. A virtual space where attorneys didn’t know the race of the jurors would prevent that from happening.
I also think it would significantly reduce the psychological strain entailed in providing in-court testimony. One of the reasons that rape prosecution are so difficult is that victims refuse to testify in court because they don’t want to be in the same room as their attacker. And with virtual space, that would lead them to feel less intimidated and nervous.
You also argue that virtual trials could include a time delay in presenting information to the jury. What’s the purpose of that?
Frequently, evidence or testimony is presented in court and then subsequently objected to. So under cross-examination, a prosecutor asks a defendant on the stand, “Were you convicted of assault earlier this year?” The defendant answers, the defense objects to that, and the trial judge instructs the jurors, I’m sustaining that objection, disregard what you just heard.
But we know from experimental evidence that jurors can’t do that. Inevitably, evidence that they’ve been told to disregard then influences their later determination. So in a controlled setting, we could institute a time delay that would prevent that evidence from ever coming before the decision makers.
In the shorter run, are there other things we could do to improve our courtrooms? Let’s start by looking at judges.
In the U.S., we think there are two kinds of judges: activist judges and umpire judges. Essentially, a person decides whether they want to be an objective, neutral judge who just calls balls and strikes, or an activist judge who follows their own agenda. But what the research from psychology says is that actually all judges are biased, and they are often biased in ways that are beyond their conscious awareness or control.
One of my favorite experiments in the book looks at whether a judge grants a person parole or not. You would think that the things that determine parole are the crime the person committed, and whether the person reformed themselves in prison.
But in fact, researchers found that the major factor was the time of day that the person came before the parole board. If you appeared first thing in the morning, you were more likely to get parole. Right before the first break in the day was the worst time. There is a clear disjunction between what the law says and what is actually determining outcomes.
I think most judges are extremely well-intentioned people who believe that they act objectively, to the best of their abilities. I think one of the best ways to change the system -- a system that we know does include quite a lot of judicial bias -- is simply to bring to judge’s attention the wealth of data that exists on what is going on. We know that African-American men receive higher bails. We know that certain people end up being sentenced to longer sentences based on demographic factors. Judges aren’t aware of those things until they look at the data.
And what about with juries?
Jurors are supposed to decide cases based on the facts and the law, and often times that’s not the case. Research shows that the jurors’ different background and experiences – what they bring to the jury panel – matter far more than differences in the legal code. It’s not supposed to matter what particular juror you happen to draw. But that is what we believe is driving a lot of outcomes in the criminal justice system.
There are actually de-biasing techniques that are being studied by psychologists as we speak. The challenge is to figure out ways to disrupt these stereotypes, by, for example, exposing people to counter-stereotypes – positive images of African-Americans, like Martin Luther King, Jr., and negative images of white Americans, like Jeffrey Dahmer. That has been shown to be effective at undermining these implicit biases.
More broadly, I think we need to think about as a society how we get rid of damaging racial associations. That has to do with depictions of African-Americans on prime-time television. It has to do with the stories that are reported on the five o’clock news. That’s where these biases come from.
Secondly, we need to think about ways to establish more diverse juries and more diverse judicial benches. The worst thing that we can possibly have is a jury or a court where all the people share the same set of biases. It’s much better if everyone is biased in different ways. Our benches and our juries are disproportionately white, male and older. That’s a problem, particularly because our legal rules themselves have been developed by white, older men over the centuries. To the extent that we cannot de-bias the population, diversity is a good second-best approach.