The New Science Behind Our 'Unfair' Criminal Justice System
TERRY GROSS, HOST:
This is FRESH AIR. I'm Terry Gross. We've all heard stories of people who served decades in prison for crimes they didn't commit. We like to think such miscarriages of justice are rare. But our guest, legal scholar Adam Benforado, says many routine procedures in the criminal justice system are bound to lead to mistakes and unfair outcomes because they rest on false assumptions about how our brains work.
In his new book, he cites research which suggests handsome defendants get lighter sentences, that parole boards are tougher when they get tired and that some common police practices encourage false identification of suspects. Benforado says there are ways to improve the system to account for unseen biases and cognitive failures that undermine the search for truth. Adam Benforado was an associate professor of law at Drexel University. He spoke with FRESH AIR contributor Dave Davies about his new book, "Unfair: The New Science Of Criminal Injustice."
DAVE DAVIES, BYLINE: Well, Adam Benforado, welcome to FRESH AIR. I'd like to start with a case that you describe in the book, a gentleman named Juan Rivera. Tell us what he was accused of.
ADAM BENFORADO: Juan Rivera was brought in for a brutal rape and murder of a young girl. And I think one of the things that really drew my attention with respect to this case was the fact that we had DNA evidence for the third trial. This man, Juan, ended up getting a third trial. That's very, very rare. And in this third trial, the reason that he got it was they finally tested the DNA evidence from the rape kit.
And what they found was Juan was not a match. And yet they tried him again, and 12 jurors convicted him again. And I thought, how could this possibly happen? And then I read a little bit more, and I saw there was a confession. And one of the things we know from psychology is that juries place great, great weight in confessions. What we also know is that confessions can be a very bad way to convict a person. Sometimes we get it wrong.
DAVIES: Well, let's talk about that. Yeah, I think a lot of jurors would just say logically, how would you possibly confess to such a brutal crime if you didn't do it? What are some of the interrogation techniques that tend to induce false confessions?
BENFORADO: It really comes down to, in my opinion, the most commonly used interrogation procedure in the United States. And this process basically breaks down the interrogation into two parts.
During the first part, the goal is simply to figure out whether this person whom you're talking to is lying. So they bring in the suspect, usually into a small, windowless room, and they ask them some provocative questions meant to reveal deceit. And the problem here is the things that detectives tend to focus on or are told to focus on, which are demeanor elements like jittery limbs or averted gaze, tend to actually be terrible ways to determine whether someone is lying. So quite frequently someone who's committed a horrible crime will look you straight in the eye and tell you that they're innocent.
Now, why this is a problem is that as soon as the detective determines that someone is lying, they shift to the second part of this commonly used approach. And the focus there is simply on gaining an admission of guilt. It's not primarily about collecting more information, revisiting that possibility that maybe this person didn't do it. It's all about getting to that admission that we need. And the techniques that are used here can be roughly summarized as sort of the classic good-cop, bad-cop routine, what's referred to as maximization and minimization. And we know from laboratory experiments that this can be highly coercive. And the people who are particularly vulnerable tend to be people with low IQs, people who are young, teenagers, people with a history of mental health problems. And Juan really was suffering from all of these issues.
And what happens is someone sitting in that interrogation seat may often think to themselves, well, this is really awful. I need to relieve this acute distress, and I need to relieve this now. I'm going to admit that I did this crime because I know that I'm innocent, and very soon after I admit, I'm sure that the police will check out my alibis. I'm sure that they will look up and find other evidence, and then I'll be released. And really this is clearly the best idea. I'll just say I did it.
And what's tragic about situations like this is that as soon as there is a confession, the police stop looking for anyone else. They don't follow up on those other leads. They don't go down these other alleys because they have a confession. We put such faith in confessions that there's no need for finding other evidence or pursuing other avenues.
DAVIES: Do we have any idea how common false confessions are?
BENFORADO: We have a sense of it. When - during the 1980s, we just started to develop the tools to actually test biological evidence and do DNA analyses. And this has allowed us a glimpse of how common these things happen. And we now have enough data that we're beginning to understand the particular factors that can increase the likelihood, things like the length of interrogation time. Juan Rivera was interrogated for hours and hours and hours. In fact, the interrogation went on so long that the detectives interrogating him got tired, and they had to bring in other detectives to continue with the interrogation.
DAVIES: One thing that a lot of people believe would help is videotaping these interrogations and certainly any which lead to a confession. Does it?
BENFORADO: I have mixed feelings about videotape evidence. Now, on the one hand, in the wake of many tragedies around the country in the last several months, there's been a real push to make sure that all officers on the streets are wearing body cameras and that squad cars all have videotapes going at all times. And I think that, in part, that's a good idea. There is good research which suggests that when people believe they are being monitored, that actually changes their behavior for the good.
On the other hand, there is research on something that's called perspective bias that gives me pause. So the particular context where this has been studied quite extensively is in the interrogation room. And the way these experiments work is they put a camera behind the person being interrogated, and they put a camera behind the interrogator. Now, if videotape presented objective reality, as we suppose that it does, that choice wouldn't matter at all. But in fact it mattered a lot in these experiments. When people watched the footage shot from the perspective of the interrogator, they tended to say, well, this looks like a completely fine, voluntary confession. But when they watch the videotape from another perspective, through the eyes essentially of the suspect, suddenly they notice all of these coercive factors. And they tended to think, well, no, actually that confession cannot come into court because it is so badly influenced by the actions of the interrogator. And so that makes me worried about this broad movement right now to switch to videotaping everything.
I think the answer and the solution is to use that videotape for certain purposes, but not for other purposes. So we might use that body camera in order to make identifications of people, but we might not do it to - as a way in an actual trial - present that evidence as a clear and unambiguous representation of what happened in the key moments.
DAVIES: There's a lot of interesting stuff here about how jurors decide who they're going to believe at trial - prosecutors, witnesses. And a lot of people would not be surprised to find that there are studies that suggest people are more likely to believe a person of their own race. There's other fascinating stuff. Are attractive people or thin people more likely to - or confident people - more likely to be believed in court?
BENFORADO: Yeah, there is evidence that a lot of physical features play a big role in whether people treats a particular witness as credible or not credible. And that's worrisome. But I think there's actually a deeper problem with jurors and that is that the things that we think are determining the outcomes of cases - that is the facts and the law - are often not what determines whether someone is convicted or not convicted, how long a sentence is. What matters most are the particular backgrounds and identities of the jurors.
So I teach criminal law. One of the areas that I teach is rape law, and my casebook takes many pages, discussing all of the different nuances across the different states. And there's a lot of emphasis on the casebook on the importance of these nuances. It really matters whether we are in a state that recognizes a defense of a reasonably mistaken belief in consent or we're in a state that doesn't recognize that particular defense. But when researchers looked into how important the law was to outcomes in, say, a date rape case, what they found was the particular legal nuances didn't matter at all. What mattered were the backgrounds and experiences of the jurors. What they refer to as cultural cognition. And these subgroups of citizens didn't break down as expected. It wasn't that men were far more likely to let the man off in a date rape scenario. It was actually within women that the most interesting break occurred. Women who were older, who were more conservative, who adhere to more traditional gender norms, were far more likely to let the man off in this particular case than women who were liberal and younger. That's a worry because a lot of what law professors do is emphasize the importance of legal doctrine. It may not be legal doctrine, though, in the criminal law sphere that's really determining the trajectory of cases.
DAVIES: One of the things we see in court is jurors trying to evaluate whether a witness is testifying truthfully. And they would look for tells, you know, whether the witness appears jittery and whether they shift their eyes a lot or doesn't make eye contact. And you write that these things - research shows these things really tell us nothing about how truthful someone's being. In fact, they can mislead us into thinking someone is being truthful when they are not and vice versa. Do the courts encourage jurors to use these, you know, supposedly common sense evaluations of the mannerisms of both defendants and witnesses?
BENFORADO: They absolutely do. And this is one of the real challenges for reform in this area is that it's not that our legal system just sits back and says nothing about human behavior. It actually weighs in on the side of myth. And so if you've ever been a juror and you are called to jury duty, you know that the starting point is this voir dire process where you're asked a bunch of questions. I was recently called onto jury, although I didn't make it ultimately onto the jury. And I was asked, you know, these questions of do you have any reason why you would be more or less likely to believe the testimony of a police officer? Now, on the jury pool that I was in, a number of people said yeah, they checked that box. The judge then came up and said, all right, well, let me explain to you what objectivity means. It means that, you know, we all have these feelings, but you've just got to put them to the side. Can you do that? Everyone in the jury pool said, yes, of course, judge I can do that. But that's not how biases work. A lot of them are not subject to introspection and control. And so it's not just that our legal system is sitting back on the sidelines. It's actively promoting false notions of human behavior, and that's really, really damaging.
DAVIES: Adam Benforado's book is "Unfair: The New Science Of Criminal Injustice." We'll talk more after a short break. This is FRESH AIR.
(SOUNDBITE OF MUSIC)
DAVIES: This is FRESH AIR, and if you're just joining us, our guest is Adam Benforado. He's an associate professor of law at Drexel University. His new book is called "Unfair: The New Science Of Criminal Injustice."
There was a study of parole boards, and it discovered that when they looked at the likelihood that the parole board would grant a petition or parole (laughter), the most important factor seemed to be what time of day his case was heard.
BENFORADO: This is, you know, one of my favorite experiments in the book. This particular study involved looking at Israeli judges, and the question was, well, what are the factors that are most important to whether someone is granted parole? I teach criminal law. I would've said - before reading this study - well, it's got to be how bad a crime the person committed and how they did in prison - did they reform or not? But when the researchers looked at it, the best predictor was simply the time of day when the particular person appeared before the panel. First thing in the morning was the best time to appear before the panel. The worst time? Right before the first break in the day. And that's worrisome if judges, right - the people that we hold out as the most objective, the most able to control their irrational impulses - are being influenced by things outside their awareness.
DAVIES: Because they were just tired.
BENFORADO: Well, we're still trying to figure out the causal mechanism. One theory is that basically to deviate from the status quo requires, essentially, mental energy and that over the course of a day, people become depleted. And the more depleted they are, the more likely they're just to stick with the status quo. And the status quo, when it comes to the parole board, is just leaving someone in prison to finish out their sentence.
DAVIES: You also cite a case of a judge, Frank Barbaro, a New York state court judge, who was tormented by a decision he'd made in a case. He was a trial judge and he found someone guilty, I believe, of manslaughter. He was a white guy who'd killed a black person. And in the end, long after the trial, he looked at it and concluded he had harshly judged the man because he'd assumed he was a white racist killing a black man - attacking a black man. When he looked at the evidence in a different way, he felt differently about it.
It's a fascinating case and it raises the question of, what do you do about judges who carry unconscious biases? I mean, the fact that they are unconscious means that it may be hard to address directly. How do you deal with that?
BENFORADO: I think the starting point is to simply raise awareness about these biases. I think a lot of judges honestly don't know about them, or they know about them but think that they involve other judges, right - that's someone else's problem, I know I'm objective.
I think when you raise awareness, suddenly judges have an incentive to start paying attention to their behavior, actually looking at what they do. I think the second step is that we need to start actually keeping better track of judicial opinions, and do that on an individualized level so that people know, as a judge am I giving black people higher bails than white people? Judges don't know that right now, and it's hard to accept the data until you look at the broad statistics.
Now, the final step, I think, is direct interventions. And we're just starting to learn when it comes to, say, implicit racial biases, ways to disrupt either the effects of these biases or to actually unsettle the underlying stereotypes upon which these biases are based. So some of that interesting research is in exposing people to counter stereotypes. So when people have looked at positive images of African-Americans, say, Martin Luther King, and negative images of white people, say, an image of Charles Manson, that helps disrupt stereotypes that may, say, link blackness with crime or violence. The challenge for researchers right now is how to use these disruptive techniques to make lasting changes to people's behavior.
DAVIES: And is there a concern that you can implant a bias by trying to counter another bias?
BENFORADO: There is some concern with that, but I think that the techniques - certainly the research on implicit racial bias has reached such a level that a lot of those concerns are being actively addressed in the research. Now, to the extent that we want to take a completely different direction - and I think, you know, this is something that we should have on the table - we could say well, you know what, we're never going to get rid of all biases. And so a next-best solution is simply ensuring that our juries and our judicial benches are diverse. The worst possible thing is when everyone is biased and everyone is biased in the same way. If you look at the makeup of our judiciary, it is primarily white, wealthy, Ivy League-educated older men. That's a problem because all of their biases are going to tend to a line. And that's going to be exacerbated by the fact that our legal system and all of our laws under our common law system have been developed by older, white, highly-educated men. I think to the extent that we can focus on term limits efforts to ensure a diverse set of judges, that's another way to get at this problem of bias.
DAVIES: Let's talk about what can be done to improve things. Let's look at interrogations. Are there some law enforcement agencies that are doing better jobs of interrogating suspects, for example?
BENFORADO: Yes, I think there definitely are. And I think the key to positive reform is casting a broad net and looking at not just what different police departments are doing in the United States, but what they're doing internationally. And one of the things that I think is a really positive development, internationally particularly, in Europe, is the development of something called the cognitive interview. And the focus of the cognitive interview - unlike the traditional approach in the United States - the cognitive interview is all about collecting information. It's not about pressuring someone into admitting guilt. It's about extracting as much information, as much data from the person you're talking with as possible. Now, what that does is, it avoids any of the problems that we know happen when we use psychological coercion. We don't really run a risk of a false confession. What we do know though is we collect all of this data which then can be compared to all of the data we have elsewhere and in which we can catch people in lies very effectively.
DAVIES: You're asking cops to forget about their hunches, it sounds like.
BENFORADO: I am asking that we rely much less on hunches, but it's not just cops that I care about. I want judges and jurors and eyewitnesses to all give up this notion that they're infallible, that their memories work the way they think their memories work, that they make good calls based on objective factors. I think we need to control for our human limitations.
GROSS: We're listening to the interview FRESH AIR contributor Dave Davies recorded with Adam Benforado, author of the new book, "Unfair: The New Science Of Criminal Injustice." They'll talk about why eyewitness testimony and eyewitness identifications can be inaccurate after we take a short break. I'm Terry Gross and this is FRESH AIR.
(SOUNDBITE OF MUSIC)
GROSS: This is FRESH AIR. I'm Terry Gross. Let's get back to the interview that FRESH AIR contributor Dave Davies recorded with Adam Benforado, author of the new book "Unfair: The New Science Of Criminal Injustice." It's about how many routine procedures in the criminal justice system are bound to lead to mistakes and unfair outcomes because they rest on false assumptions about perception and how our brains work. Benforado is an associate professor of law at Drexel University.
DAVIES: Let's talk about eyewitness testimony and eyewitness identifications. One of the most effective things in a court is for a witness to look in the courtroom and say, yep, that guy did it. I recognize him. How often are eyewitness identifications inaccurate?
BENFORADO: The way a normal identification procedure works is the police have someone in mind, a suspect, and they bring that person in. And then they pull in four or five other people to stand next to that person as the eyewitness looks at the lineup. Now, we know a third of the time in real eyewitness identifications, the eyewitness picks out one of those fillers.
DAVIES: And what are some of the factors that lead to misidentification?
BENFORADO: There are a set of factors that are outside the control of the police departments. So we know, for instance, that cross-race identifications are a lot more likely to be inaccurate. We know that people are a lot less accurate if they have seen someone at dusk rather than at midday. We know that people tend to be less accurate when they physically exerted themselves at that moment of memory encoding.
Now, those are all factors that are out of the control of the police department. But there are a whole host of issues that are raised by how police officers handle eyewitnesses. And one of the things that I think is a very important step for us with respect to eyewitnesses is treating memory more like we treat other types of trace evidence. If you think about a bloodstained handkerchief - how careful we are to preserve that evidence, how careful we are to track sort of how it's handled over time and who has it at any given moment, we don't take any of those similar precautions with memory.
And yet what we know from the psychological evidence is memory is very, very fragile and very malleable, just like blood evidence is. So very simple things could be enacted today which would make a big difference. One of the things is simply ensuring that the person administering the identification procedure doesn't know who the suspect is. It's clear from both laboratory research and real-world evidence that when the police officer who is handling the eyewitness knows who the person is that the police have their eye on, they can vary subtly sway the person, often without even being aware they're doing that. So imagine that the eyewitness starts to pick out one of those fillers. Well, the police officer may just say, now, now, ma'am, take your time. We're in no rush here. The person - the police officer doesn't think, OK, well, I'm tainting this eyewitness identification. After all, that seems a good advice. There's no rush. Make sure you're certain. There's a lot of consequences here. But that kind of subtle behavior actually does have an impact.
DAVIES: What are other ways that a lineup can go wrong?
BENFORADO: Another great example of something that many police departments do, which is very problematic, is using multiple eyewitness identification procedures. So imagine that there's a robbery. They may bring one of the eyewitnesses from the bank, drive them by in the squad car and say, hey, do you see the person who robbed the bank anywhere on that street corner. Well, there's someone with handcuffs sitting there, and the eyewitness may say, yes, that person.
Now, two weeks later, that same eye witness is brought in and shown a lineup. Now, the question is are they remembering the person they saw sitting on the stoop, or are they remembering the person from the bank? We don't know. But imagine that they again point that person out. They choose the same person. This time, they say, I'm a hundred percent certain.
Then, very commonly, that particular eyewitness is then brought into court many months later to do an in-person identification. And at that point, the person says, yes, I am as certain as I have ever been on anything. And juries, we know, find this incredibly compelling. After all, the person has identified the same person three times and is absolutely certain.
DAVIES: So what happens is the memory becomes more reinforced with each successive iteration.
BENFORADO: Absolutely. And that first memory may not be a true memory at all. That was a highly suggestive procedure that was used. After all, there's a person sitting handcuffed on the side of the road. That really suggests that the police must know that this person is guilty.
DAVIES: You write a lot that jurors often make decisions and trials which are influenced by things that shouldn't matter, like how confident a lawyer is or by an expert witness who really might not have anything that meaningful to say but which can impress the jury or in some cases, you know, thinking that they can tell when someone's telling the truth. And you have a rather radical idea to overcome all of this - eliminating live trials. Explain.
BENFORADO: I think that we need to always have our eye on what's our legal system going to look like 10 years, 50 years, 100 years out. And one of the things that I think we ought to be thinking about is better controlling of how trials run. And one of the ideas that I suggest in the book is the possibility of virtual trials in which a lot of the things that we know lead to wrongful convictions and incorrect outcomes could actually be controlled.
So it shouldn't matter how attractive the witness is. It shouldn't matter whether the attorney is wearing a particular tie or gesturing in a certain way. The bombastic exchanges which occur in many trials - those shouldn't determine the outcome at all. That makes for good television but terrible justice. Now, if we controlled the situation of the courtroom, if we used, say, avatars and used a time delay, we could remove a lot of these biasing factors and ensure that jurors were focused on what we purport is the most important thing to the outcomes of trial.
DAVIES: Avatars meaning that - would be an avatar for every witness, an avatar for the attorneys, an avatar for the judge so that the words would come out in some sort of neutral way that the juror could evaluate. And you said a time delay? What about a - why a time delay?
BENFORADO: So one of the things we know about evidence is frequently, evidence which is actually not admissible that is misleading or that is irrelevant actually appears before jurors. And the way that we deal with it currently is the judge then instruct the jury, disregard what you just heard. And we treat that as if that's effective. Research suggests that's terribly ineffective. In fact, judges themselves are unable to prevent themselves from being influenced by such evidence. So the idea of a time delay would ensure that the decision-makers never ever hear about that objectionable evidence.
DAVIES: You know, I can hear a police detective listening to this or a prosecutor or an attorney or a judge saying, yeah, this is some egghead law professor who's read a bunch of studies. But the reality is that everybody is doing their best. And the system does get a lot right and that you don't have - you don't know what it's really like out there on the street, in the interrogation rooms, in the courts and that if you had more practical experience, you would see that it's not as bad as you think and that there are reasons for the way people behave as they do. Do you feel like you've got enough of a feel for the system to be judging it?
BENFORADO: I think absolutely. And I think there are going to be some people who are deeply skeptical. Obviously, we've been doing this in these ways that I suggested leading to deep problems for a very long time. There are a lot of vested interests - not just economic interests - in keeping things the way they are. Now, I will tell you this book is been out less than a week. I am not a famous person. I have been receiving e-mails from judges, from police officers telling me in particular after I wrote a New York Times op-ed, yes, this is what people need to hear.
I just heard about two days ago from an officer in New Jersey who said, you know what? I have been focused on evidence -based policing for a very long time. It's been an uphill battle. But with the amazing Internet, I have recently connected with a whole bunch of other police officers who are very interested in collecting data, in running experiments to figure out what's effective when it comes to interrogating people. What's effective when it comes to handling eyewitnesses? And I just want you to know we have started a new society focused on evidence-based policing, and we are committed to reaching across the United States.
I know that there are prosecutors out there who are starting programs based on making sure that the role in their office of the prosecutors is not just gaining convictions but looking for mistakes, ensuring justice, making sure that innocent people are never put away. Now, are there other prosecutors' offices who are going to listen to this program and say, you don't know anything, the adversarial process works? Yes. But I want to convince people that the innovators out there, the jurisdictions that are trying new things have it right. And they are going to be more successful. And they are going to be delivering justice to the communities in which they inhabit.
DAVIES: Well, Adam Benforado, thanks so much for spending some time with us.
BENFORADO: It was an absolute pleasure.
GROSS: Adam Benforado spoke with FRESH AIR contributor Dave Davies. Benforado's new book is called "Unfair: The New Science Of Criminal Injustice." Our book critic Maureen Corrigan has four suspense novels to recommend for summer reading after we take a short break. This is FRESH AIR. Transcript provided by NPR, Copyright NPR.