I’ve talked about the Clarence Elkins case quite a bit already, and his case is a perfect illustration of tunnel vision. You’ll recall that Elkins was wrongfully convicted of raping and murdering his mother-in-law and raping his six-year-old niece during the same attack. Immediately after the attack, the niece said that the perpetrator “looked like” her Uncle Clarence. When authority figures, including the police, talked to her, however, her story morphed into “it was my Uncle Clarence.” But it was dark in the house where she and her grandmother had been attacked in the middle of the night, and the little girl had been knocked unconscious at the beginning of the assault. So she, of course, could not have gotten a very good look at the perpetrator. But despite this weak evidence, the police arrested Clarence at his home within hours of the crime and then assured the public that they had the dangerous criminal in custody. They patted themselves on the back for acting swiftly to protect the public.
From that point on, nearly every piece of evidence that was discovered during the investigation decisively pointed to Elkins’s innocence. But the police routinely ignored all of it. Indeed, after the police had arrested Elkins and announced they had solved the crime, there was no going back. Their minds were made up. Tunnel vision had set in.
That evidence of innocence included the fact that Elkins’s then-wife, Melinda, told the police that Clarence had been asleep in their bed at the time of the crime. One of their children had been awake all night sick, and Melinda had to stay up to take care of him. There is no way, she told police, that Clarence could have gotten out of bed, driven thirty minutes to his mother-in-law’s house, committed a murder and two rapes, and driven thirty minutes back and gone back to bed without her knowing it. And he had no motive. There was some friction at times between Clarence and his mother-in-law but nothing out of the ordinary for a relationship of that nature.
Melinda said there was no way she would lie—after all, it was her own mother who had been murdered and her own sister’s daughter who had been beaten and raped. But because of the police’s tunnel vision, they knew that she was lying. She was labeled as a “stand by your man” countrywoman who would rather protect a murderous husband than her own mother and niece. The prosecutors eventually pressed that same theory to the jury when Melinda testified in Clarence’s defense; in essence, they suggested to the jury that she should not be believed.
Yet every piece of forensic analysis failed to connect Elkins to the crime scene. The scene was a bloodbath, with bloody palm prints, hairs, and all sorts of evidence on the floors and walls. None of it matched Elkins. Not a shred. No matter, the police and prosecutors said. The mother-in-law’s home was very dirty, they claimed, and she must not have cleaned it in ages. So the fingerprints and hairs and other such things must have been left by prior visitors. Somehow, they claimed, Elkins was able to commit this crime without leaving so much as a trace behind. And though the perpetrator would have no doubt been drenched in blood, they could find no blood connecting Clarence to the crime in his car, on any of his clothes, or anywhere in his house. The police even had the drains of Clarence’s house scoured for blood, assuming Clarence showered the blood from his body. But their search turned up nothing.
At this point, an objective observer might think that the testimony of a six-year-old girl who witnessed her attacker for a few seconds in the dark, under great stress, was undermined by Elkins’s solid alibi and the lack of forensic evidence connecting him to the scene. Indeed, it would seem unimaginable to any objective observer that Elkins could have committed such a bloody crime and left no evidence at the scene, while removing all traces of blood from his car, clothing, and house in just a few hours. Yet, because the police and prosecutors suffered from tunnel vision, they continued to push forward. And they did so with ferocity—seeking the death penalty against Clarence.
Based on the testimony of the child, Elkins was convicted. Thankfully, the jury rejected the death penalty but he was sentenced to life in prison.
After Elkins had served seven and half years, DNA testing exonerated him, as I’ve said already, and proved that Earl Mann was the true perpetrator. Earl Mann is now serving life in prison for the crimes for which Clarence Elkins had been originally convicted.
So, in addition to ignoring all the evidence that pointed to Elkins’s innocence, police and prosecutors also ignored the fact that a violent child rapist had just been released from prison and was living two doors down from the crime scene. That should have been rather glaring. Earl Mann even looked quite a bit like Clarence Elkins. But it gets even worse. It was later discovered in Elkins’s civil suit against the police department that before Elkins’s trial, Earl Mann had been picked up by the same police department for a violent robbery and had said something to the effect of, “Why aren’t you arresting me for the murder of Judy Johnson?” Judy Johnson was Elkins’s mother-in-law. That comment was obviously a thunderbolt, and the arresting officer informed the detective bureau investigating the murder about it. But the detectives were so beholden to their theory of the case, and to Elkins’s guilt, that this incriminating comment was never followed up on. It was just ignored. Despite the fact that Mann happened to live two doors down from the crime scene, happened to resemble Clarence Elkins, and was known to the police as a violent, serial offender. And the police report documenting Mann’s statement wasn’t even disclosed to Elkins’s defense team prior to trial!
Thanks to appalling tunnel vision, Elkins spent seven and a half years in hell, and Earl Mann went on to rape and physically abuse several other children before he was eventually caught and imprisoned for life.
The Elkins story might seem bizarre, or almost unbelievable, to a person not familiar with the criminal justice system. But it is not atypical at all. Tunnel vision is rampant in our system.
• • •
Tunnel vision occurs when we develop an initial belief or suspicion, become wedded to that belief, and then interpret or even twist all subsequent information we encounter in order to confirm it. It’s a common human tendency that arises in a variety of situations in our lives, because it is “partly innate” and “part of our psychological makeup,” as Keith Findley has observed.1
Tunnel vision at first glance appears to be similar to confirmation bias. But it’s more than that. All of the psychological problems discussed in this book work together to create tunnel vision. Because of confirmation bias, we see new information through the warped lens of our initial belief, which we are loath to let go of. “People tend to see information that confirms their hypothesis and to avoid information that would disconfirm their hypothesis,” Findley writes.2
And given our malleable memories, we recall past events with memories that we have reshaped to support our new belief. This sometimes results in the “knew-it-all-along effect” or “hindsight bias,” or the tendency to see events as more predictable than they actually were. Cognitive research of memory has clearly demonstrated our tendency to reconstruct our memories of past events to be consistent with our later beliefs.3
Moreover, because humans are not good lie detectors, but rather not much better than a coin flip at determining who is telling the truth and who is lying, tunnel vision causes us to selectively make these determinations based on which witness accounts fit our hypothesis of the crime.
When external political pressure is applied to reach a certain conclusion consistent with an initial hypothesis, such as the pressure on police and prosecutors to solve a horrific crime, all of these problems are greatly exacerbated, and tunnel vision can wreak havoc.4
Tunnel vision served an important purpose in bygone eras. A tribal leader in charge of a hungry tribe needed to be decisive, and act quickly in interpreting various clues, to determine which direction the game herd likely moved in overnight so that his tribe could mobilize in time to keep pace with the herd. A deep thinker, on the other hand, who might take all day to reach a conclusion after evaluating all the available clues, might pick the right direction a higher percentage of the time than the decisive leader, but his or her tribe might starve, because such a deliberate, careful process would allow the game herd to advance beyond reach.
Evolution thus favored quick decisions and the ability to ignore distractions while remaining wedded to the most obvious option. As a result, our brains innately engage in what are called “heuristics”—hardwired mental shortcuts that help us to make decisions quickly—jumping to conclusions, one could even say, without getting bogged down in too many distracting details.5 But psychologists have realized that while heuristics were necessary in past eras, and can be helpful in many aspects of life today, they can sometimes lead to disastrous results in our complex world. And in the criminal justice system, our innate psychological instincts can cause serious problems if we’re not aware of them and don’t try to keep them in check.
Tunnel vision is a powerful force. Political leaders, for example, frequently fall prey to it in matters of foreign policy. George W. Bush and his administration developed suspicions, if not paranoia, after 9/11 that Iraq and its leader, Saddam Hussein, were hiding weapons of mass destruction. Leading up to the U.S. invasion of Iraq in 2003, they repeatedly chose to believe unreliable sources, such as informants who had been “spoon-fed” to tell the administration what it wanted to hear, and forged and other unreliable documents on this matter. Meanwhile they discounted hard data and information from more reliable sources that indicated that Iraq had already destroyed its WMDs after the earlier Gulf War.6 After the invasion commenced and it was discovered that Iraq no longer possessed WMDs, a governmental commission assigned to investigate the fiasco reported to the president:
In the case of Iraq, collectors of intelligence absorbed the prevailing analytic consensus and tended to reject or ignore contrary information. The result was “tunnel vision” focusing on the intelligence community’s existing assumptions. . . . For example, several human sources asserted before the war that Iraq did not retain any WMDs. . . . But the pervasive influence of the conventional wisdom—that Iraq had WMDs and was actively hiding it from inspectors—created a kind of intellectual “tunnel vision” that caused officers to believe that information contradicting the conventional wisdom was disinformation. Potential sources for alternative views were denigrated or not pursued by collectors.7
In a nutshell, the Bush administration had become “too wedded” to its initial assumptions and then failed to analyze all new evidence objectively, resulting in an expensive and deadly war that disproved the very assumptions that the administration had so stubbornly refused to question.8
But George W. Bush is certainly not the only U.S. president to suffer from tunnel vision. Lyndon B. Johnson repeatedly ignored data and advice from his advisors that the war in Vietnam was unwinnable.9 George Reedy, an advisor to Johnson, said that President Johnson “had a remarkable capacity to convince himself that he held the principles he should hold at any time, and there was something charming about the air of injured innocence with which he would treat anyone who brought forth evidence that he had held other views in the past. It was not an act. . . . He had a fantastic capacity to persuade himself that the ‘truth’ which was convenient for the present was the truth and anything that conflicted with it was the prevarication of the enemies. He literally willed what was in his mind to become reality.”10 Most if not all of us are guilty of similar self-delusions in our lives.
In 1962, John F. Kennedy and his administration clung to the belief that the Soviet Union would not place missiles in Cuba, and were very slow and reluctant to accept any contrary information, although it was mounting. In 1968, Johnson’s administration discounted clear and reliable evidence that the Soviet Union was preparing to invade Czechoslovakia, because the evidence conflicted with officials’ preconceived beliefs, until the invasion happened. Five years later, the Nixon administration could not be shaken from its belief that Egypt and Syria would not attack Israel, despite a “mountain of evidence” to the contrary.11 A survey of nearly any past presidential administration would find similar examples of tunnel vision, often with disastrous results. But it’s not just presidents who suffer from tunnel vision. Everyday people experience it in their daily lives, at work and at home.
Even love, for example, can be a strong facilitator of tunnel vision. Those who have been in troubled relationships may relate to this. Indeed, I know this from personal experience. In 2004, when my wife at the time first told me she wanted a divorce, I was devastated. We had had a near-perfect relationship for more than fifteen years, I believed. I pushed back against the divorce for more than two years, confused as to why it was happening and in anguish. After we were divorced in 2006, I was depressed for months, convinced my world was rocked forever. Things would never be the same. What a waste, I thought, to throw away such a wonderful relationship that had produced two beautiful, young children. And to put them through a painful divorce for no good reason seemed callous and incredibly unnecessary.
I had grown up in a household where divorce was greatly frowned upon. My parents enjoyed a storybook marriage filled with daily expressions of love, joy, and support, with no visible tension between them. Growing up, I knew I would have that too. I knew it because I didn’t know anything else. My parents’ relationship was my only example, and I naïvely thought that was just how it was with any couple that made a daily effort to love and support each other. It was easy. So a marriage like my parents’ became an ingrained part of the vision I had for my life.
I was in love at the beginning of my relationship with my ex-wife, and saw growing old with her in happiness and joy, like my parents did with each other, as my only path in life. But as time passed after our divorce, I could feel the layers of tunnel vision that I had created during our marriage begin to peel off one by one. I began seeing our relationship in a different light, very gradually, step by step. I began to comprehend problems in our relationship, which I had simply been unable to see. I eventually realized that my ex-wife and I were quite incompatible and just weren’t right together. She had been right all along.
I had not been happy in the marriage either, or at least wouldn’t have been if I had been honest with myself, but had convinced myself that I was happy through self-rationalization about the problems and copious amounts of revisionist history. What had actually devolved into an unhappy and unsatisfying relationship remained a near-perfect union to me, given the way I had limited my field of vision. By the end of 2007, however, after enough time had passed, I was able to see more clearly, to shake myself loose from my self-delusions, and to feel thankful toward my ex-wife for having the courage to break things off despite my naïve and stubborn inability to see what she had realized long before I did—that it wasn’t anyone’s fault, but we just weren’t right together.
I am, of course, not the only person who has suffered from tunnel vision in matters of troubled relationships. Although I didn’t mention the role of tunnel vision in relationships in my short proposal for this book, a friend who read the proposal said, “You know, you should talk about how love can cause tunnel vision in bad relationships. I was in a terrible relationship for years. He was mean. He didn’t have a job. Nothing about him was good, but I was somehow blinded. I was a wreck when he broke up with me. Everyone could see the problems but me. I could only see it much later. I had tunnel vision.” And so it is with many in love, I believe, particularly those in unhealthy relationships. As Benjamin Franklin once quipped, couples in troubled relationships tend to keep their eyes “wide open” before marriage, and then “half-shut afterward.”
But it’s not just one’s field of vision that becomes narrowed. Couples in rocky relationships constantly revise their memories to comport with their current perception of their relationships. As one psychologist has written, “distortions of past events—or complete amnesia—kick in to confirm” each partner’s beliefs.12 If a partner is like I was in my first marriage, and desperately wants to believe that they are in a good marriage, they will revise their memory of past events in a glowingly positive light to maintain that belief. But if a partner has decided that the marriage is over, and wants a divorce, they will tend to recast past events in an unfairly negative light. One marriage psychologist reports that when she asked a couple she was counseling how they met, the wife responded, “At school, where I mistakenly thought he was smart.” The wife went on to say that she didn’t make a mistake in choosing her husband; rather, he made a mistake by tricking her to make her think he was intelligent.13 Alas, marriage counselors must be great repositories for twisted and revised memories.
• • •
In like fashion, police and prosecutors, because they are human, are prone to fall in love with their initial suspicions and beliefs. Criminal investigation is a highly competitive and stressful enterprise, and police and prosecutors feel both internal and external pressures to solve cases on their docket. The inability to solve a horrendous crime is emotionally stressful and nerve-wracking, and can lead to low self-esteem and negative repercussions in the workplace. Indeed, police and prosecutors are often evaluated on their “clearance rates”—the percentage of cases assigned to them that they are able to solve and that result in convictions.14
They are not emotionless robots. They have to visit bloody crime scenes, and hold the hands of victims’ families. They become personally sickened by violent crimes, because they are immersed in the horror, and become deeply emotionally invested in obtaining justice for the victims. As a result, nothing feels better than that initial “Ah-hah! This is very interesting. . . . Here we go!” feeling when a suspicious fact surfaces that finally brings a suspect into focus. Police and prosecutors, particularly with heinous or high-profile crimes, are desperate for those initial suspicions to be correct, because otherwise they must start over from ground zero and deliver more bad news to the heartbroken victims or victims’ families.
I know this from my own experience as a prosecutor. After that initial Eureka! moment when a solid suspect jumps to the forefront, your mind runs through all the old evidence in the case over and over again with great excitement. Facts that were meaningless before suddenly take on a sinister light now that you have a suspect. “That explains why he called the victim thirty minutes before she was murdered,” you think. “He must have been making sure she was home by herself!” “Now it makes sense,” you realize. The pieces of the puzzle start fitting together. The momentum builds. Your mental adrenaline kicks in. You get more and more excited with each new “realization” as you replay the case in your head. You try to make each piece of evidence fit your new theory.
A witness you interviewed who provided the suspect with an alibi was obviously lying, you now believe. You look for reasons not to believe him. “He’s the suspect’s friend, so of course he’s lying to protect him. It makes sense now.” Or: “Did you see how much he fidgeted and avoided eye contact with me? He seemed nervous like he was hiding something.” You suddenly convince yourself that you never believed this witness, even if, in fact, you had—you knew something was fishy with him from the beginning. You knew it all along.
But the witness who provided damning evidence against the suspect was telling the truth, you now believe. And you convince yourself that you knew that all along, too. With each new rationalization, you become more and more entrenched in your belief.
I did this as a prosecutor. I did it every single day. And I did it honestly. I was completely unaware of confirmation bias. I didn’t think my memory was malleable. I thought I could truly tell when witnesses were telling the truth or not, because I was a good lie detector. But I now know, as clear as day, that this isn’t true. And I wouldn’t have been able to admit to myself at the time that institutional political pressures had any bearing whatsoever on my decision making.
But tunnel vision was part of the culture of my prosecutor’s office. We shared our theories with those around us. Everyone took part in bouncing their ideas off their colleagues, and coming up with ways to make the pieces fit together. The excitement of “cracking the case” and coming up with ways to make all the pieces fit was part of the fun of the office. And as each new piece was made to fit, we became more beholden to our hypothesis. We fed off each other. To the extent that we tried to see arguments on the other side, it was just so that we could be prepared to shoot them down at trial.
• • •
One of my first cases as a prosecutor illustrates how tunnel vision works. In the story that follows, I have changed the names of the participants and some less significant details for the sake of anonymity. The defendant, whom I’ll call Diego Miranda, had been a low-level equipment manager for a professional sports team in the New York City area. The victim, whom I’ll call Jackie, was a famous, all-star professional athlete for the same team. Although Jackie was a millionaire, he was young and disorganized and had a tendency to overspend. Jackie’s accountant tried to rein in in his spending, so each month he sat down with Jackie and went over Jackie’s returned checks to see what he had purchased and to lecture him about living within his means. At one such meeting, Jackie denied having written some of the larger checks, most of which had been written to pay off credit cards. Jackie noted that the checks in question had credit card account numbers written in the memo section, and those numbers didn’t match any of Jackie’s credit cards.
When the accountant asked him who had access to his checkbook and could have written them, Jackie named Diego Miranda. Jackie explained that Diego had been hanging around his house along with other “wannabes”—people who try to ingratiate themselves into the social circles of professional athletes. Jackie said Diego could have gotten his hands on one of his checkbooks, which Jackie just left sitting around, and that he recognized Diego’s handwriting on some of the checks. Diego had obviously forged his signature, Jackie claimed, and used the checks to pay off his credit cards. So the accountant brought the checks to us and the investigation started.
When he brought the checks to us, the accountant told us that he had already confronted Diego about the checks, and Diego broke down and cried, apologizing for taking Jackie’s money. So my office had Diego arrested.
Diego’s attorney insisted, however, that Diego was innocent, and he asked to meet with me. At the meeting, he told me that Diego had worked for Jackie. Jackie had started using Diego as a personal assistant, having Diego run errands, purchase things Jackie needed like clothes and electronics, pick up his dry cleaning, get his car washed, make sure there was always food and beer in Jackie’s refrigerator after games, and other things of that nature. The relationship escalated to the point where Diego was spending most of his time when not working for the team performing personal concierge services for Jackie.
Jackie had given Diego a checkbook to pay for some of the things Diego purchased for Jackie. But Jackie was disorganized, the attorney said. He rarely had cash on him, and he kept forgetting to pay Diego for all his work. When Diego asked Jackie to formally hire him so that he could receive a regular paycheck, Jackie told him that his accountant would never agree to such an arrangement, so it had to stay “off the books” and remain a secret.
Jackie told Diego, the attorney claimed, “Go ahead and buy some stuff to make up for what I owe you, and then buy yourself some stuff each month going forward as your salary. Put the things you buy for yourself on a credit card, and then use my checks to pay off your credit cards. That way my accountant will never know—he’ll just think I wrote those checks to pay off my own credit cards.” Diego’s attorney said that even though such an arrangement would sound strange to a normal person, Jackie was a young millionaire who was disorganized and immature, and flew by the seat of his pants. Diego’s attorney also told me that shortly before Jackie made the theft allegations against Diego, Jackie and Diego had had a falling out over Diego’s interactions with a woman Jackie was dating, and that Diego had been fired from his under-the-table gofer job. He also claimed that there was a misunderstanding about Diego’s alleged “confession” to Jackie’s accountant. Jackie had told Diego that their business relationship had to remain a secret because the accountant would never approve of it. So when the accountant confronted him about the checks, Diego panicked, not knowing whether, if he revealed the secret, it would get Jackie in serious trouble, undermining a relationship important to Jackie that could hurt his chances for endorsements or God knows what else. So, with little time to think, Diego confessed, figuring that once he told Jackie what had happened, Jackie would best know how to fix the situation. But afterward, because of their falling out or because of Jackie’s fear of the accountant knowing the truth, Jackie stabbed Diego in the back. Diego was Jackie’s fall guy, he claimed.
After hearing this story, I interviewed Jackie. Jackie stated that he had never hired Diego as a personal assistant. Diego, Jackie claimed, was a groupie or “hanger-on” who tried to suck up to celebrity athletes. Diego was always showing up at his house and hanging around at parties, so all the professional athletes just used him to make beer runs or drive their girlfriends home from clubs—things of that nature. But they didn’t pay Diego, other than picking up his tab for food and drinks when they were out on the town together. Rather, they just took advantage of him because he was one of the many eager fans who would do anything to hang around the professional athletes after hours.
I went forward with the case. My office had already brought charges against Diego. Jackie’s story made sense, and Diego’s story seemed far-fetched. After interviewing Jackie and watching his demeanor, I believed him. I discounted Diego’s story as told to me by his lawyer. And the amount that appeared to have been stolen, around $100,000, seemed too high to constitute a salary for a mere gofer position, even for a multimillionaire like Jackie.
A day or two before the jury trial started, I finally received all of Jackie’s bank records for the months around the time the alleged thefts had occurred. To my shock, there were checks in what I recognized as Diego’s handwriting in small amounts written to businesses like dry cleaners, car washes, and grocery stores. Various checks of this nature, all in what I recognized as Diego’s handwriting, with Jackie’s signature as it looked when forged by Diego. When Diego’s attorney received his copies of the checks, he called and said, “See, this is what I was talking about. Diego was picking up Jackie’s dry cleaning, getting his groceries, and making a lot of Jackie’s day-to-day purchases. He was Jackie’s personal assistant. The remaining checks you’re bringing charges on were written by Diego to pay off Diego’s own credit card expenditures, but they were authorized by Jackie as Diego’s payment for all his work. They had a bizarre setup, but that’s Jackie for you.”
My trial was starting in the next day or two, and I felt like I had been kicked in the stomach. These new checks seemed to support Diego’s story, and contradicted what Jackie had told me, but I just somehow knew it couldn’t be true. I was positive that I was right. By this time, I was deeply wedded to the notion that Diego was guilty. So I thought and thought. And then it finally dawned on me: Diego was smarter than I had given him credit for. He stole Jackie’s checkbook, but rather than start immediately writing big checks to pay off his credit card debt, he devised a scheme whereby he would first use the checks to pay for dry cleaning or buy groceries—small, inexpensive purchases like that. These were “test checks” to see if the forged signatures would make it past the bank employees. And since some were for businesses near Jackie’s house, everything would seem copacetic to Jackie’s accountant. These “test checks” were designed to get the bank employees and Jackie’s accountant used to seeing checks written in Diego’s handwriting—to condition them to breeze past checks with Diego’s handwriting. And if they did notice the handwriting and Diego got caught, it would only be for very small amounts—no big deal for a millionaire like Jackie. Nothing would come of it. But if time went by and no alarm bells sounded, then Diego would know it was smooth sailing, and he could start writing bigger checks to pay off his own credit cards.
As soon as this light bulb went on in my head, I knew I was right. I ran my theory past my colleagues, and they agreed with me. “That’s awesome. That makes sense,” they told me. And I was proud of myself for rescuing the case from the brink of disaster.
At trial, I presented to the jury my theory of Diego’s brilliant scheme of writing test checks in smaller amounts, and then when they went undetected by the bank and Jackie’s accountant, writing bigger checks to pay off his credit cards. Diego testified to his story that he worked for Jackie, and that the big checks written to Diego’s credit cards were his salary, authorized by Jackie. Diego suggested that Jackie was lying because his accountant had busted him again for overspending, and because Jackie had a vendetta against Diego due to their falling out. Jackie denied all of this and said it was theft. I told the jury something like, “Why would a professional athlete like Jackie lie about this? He has so much to lose. He would have to be evil to frame a low-level employee that worked for his team. There’s no evidence that he’s evil.” The jury convicted and Diego was sent to federal prison.
Although I believed Diego was guilty at the time, today I can no longer be completely sure. I look back on that case now, like most other cases I prosecuted, and I see that I approached the case with the same sort of tunnel vision that I now see from prosecutors and police in my innocence cases. Although I didn’t know what really happened, I was convinced that I did. I came up with the theory about “test checks” out of the blue the day or two before trial, even though at first glance the checks seemed to support Diego’s story. I did it because I was prone to twist and turn any new evidence I received to make it fit my preconceived theory, to which I had been wedded for weeks. I had worked so hard on the case, having for weeks believed that Diego was a liar, that the only option I could consider when the new checks arrived the day before trial was that they had to be part of some devious scheme by Diego. So I made up a theory that fit my preconceived beliefs and pushed forward as if my theory amounted to proof beyond a reasonable doubt.
Although I believed Diego was guilty at the time, today I have no idea whether Diego devised a brilliant “test check” scheme, or whether he worked as a personal concierge for Jackie as he claimed. My theory might have been correct, but I don’t know. I have no idea whether Jackie framed Diego because he was mad at him, or because he didn’t want to get in trouble with his accountant again for overspending, or both. But looking back on it, I do know that I was supposed to present proof beyond a reasonable doubt of Diego’s guilt, and instead I presented a theory that I made up out of the blue right before trial. I presented it with passion because at the time I knew that I was right, and firmly believed that Diego Miranda deserved to go to prison.
The criminal justice system believes this sort of ambiguity is perfectly fine, because the jury can look into the eyes of the witnesses and tell who is telling the truth and who is lying. If Diego were telling the truth and Jackie were lying, the jury would certainly know it. I believed that back then too. I believed Jackie was telling the truth and Diego was lying, and that the jury would be able to see the truth as I did. But I now know that none of us are accurate human lie detectors, and that our lie-detecting abilities are even worse when we are deeply committed to a theory. I was more apt to believe Jackie and disbelieve Diego simply because I was deeply invested in Diego’s guilt. And so was the jury. As we saw earlier, despite our lofty “presumption of innocence,” jurors tend to believe that the police and prosecutors wouldn’t waste their time unless the defendant were actually guilty. Rather than presuming innocence, they tend to start off with a predisposition against the criminal defendant and in favor of the police and prosecution.
My tunnel vision was amplified in the case because it was high-profile, given Jackie’s involvement along with that of several other famous professional athletes and coaches who testified. There was even a famous rapper slated to testify, who had been present at some of Jackie’s parties when Diego was there. I knew the case was going to be covered by ESPN and the New York Times, among other media outlets. A week before trial, I had already arranged with a courtroom sketch artist, who had been contracted by the media for the case, to buy his original sketches after the trial to frame for my office wall. I still have two of those sketches today. One, which depicts me asking Jackie questions on the witness stand, hangs in my parents’ home. With all the work I had invested in the case, and how my reputation was at stake, and all the attention the case was getting, I probably would have taken any new evidence of innocence and come up with any theory necessary to keep the case on track. Of course, with any serious case—like murder or rape—the stakes are naturally very high and the pressure intense, regardless of media attention.
Prosecutors obtain convictions like I did in Diego Miranda’s case every single day. It’s the norm in hotly contested, close cases. Prosecutors develop theories, or ways to spin the evidence, that are nothing more than speculation churned in a tub of biased, one-sided tunnel vision. They then confidently present these theories to the jury as fact—as “proof beyond a reasonable doubt.” The system then relies on the jurors as human lie detectors to sort out the mess and determine which side’s theory is right, with many jurors tending to start with a preexisting bias in favor of the prosecution. And with the defense operating at a serious disadvantage in terms of resources, often without investigators or experts to back up its competing theories. That’s just how our adversarial system works.
But it’s important to recognize that this is far from an error-free process. It’s actually quite flimsy and subjective, to be honest. It may be that we have no other option in hotly contested, close cases like Diego Miranda’s than to just throw the evidence to the jury and let them decide. We can’t just stop prosecuting crime because we haven’t figured out a foolproof way to divine truth. But we need to at least call a horse a horse and recognize our system’s unstable foundations, built on human beings and all the flaws they bring with them. Because of human limitations, we don’t have a foolproof way of finding the truth—of reconstructing past events with a high level of accuracy. Instead of prosecutors presenting their made-up theories to the jury as unassailable facts, and then trying to convince the jury that the prosecutors actually know what happened, perhaps they should be honest and call their theories what they are—theories. Then let the jury decide between the various possibilities in a setting more conducive to objective evaluation. After all, it’s a prosecutor’s job to seek justice, not to win cases.
And as things currently operate in our world of unbridled tunnel vision, we treat each conviction as the “gospel,” as if there’s no chance juries ever get it wrong. So when new evidence comes to light years later suggesting that the person convicted and sent to prison might have been innocent all along, we pretend that the original conviction was the word of God, which can never be challenged. “How dare you argue that this guy could be innocent?” prosecutors often ask. “He was convicted by a jury of his peers. He’s obviously guilty.”
• • •
Michael Morton of Williamson County, Texas, a case discussed earlier, was also a victim of tunnel vision. On the morning of August 13, 1986, Michael woke up a little miffed at his wife, Christine. The couple had gone out to dinner the evening before for Michael’s birthday. When they got home, Michael was hoping for birthday sex and put on an adult film to watch with his wife. But Christine promptly fell asleep, leaving Michael upset and frustrated. Before leaving for work the next morning, Michael wrote Christine the following note: “Chris, I know you didn’t mean to, but you made me feel really unwanted last night. After a good meal, we came home, you binged on the rest of the cookies. . . . I’m not mad or expecting a big production. I just wanted you to know how I feel without us getting into another fight about sex. Just think how you might have felt if you were left hanging on your birthday. I L Y—M [I love you—Michael].” Michael left the note on his wife’s vanity and left for work. By the time he returned home later that day, he was a murder suspect.
The Mortons’ friend and neighbor Elizabeth Gee discovered Christine’s body. She’d entered the Morton household after finding the couple’s two-year-old son wandering around outside wearing only a diaper. Upon entering the couple’s bedroom, Gee found Christine bludgeoned to death on sheets stained in semen.15
Police arrived at the home before Michael was aware of what had happened. They immediately found Michael’s note on his wife’s vanity. The note fueled the police’s inherent suspicion of Michael as the victim’s spouse, leading them to jump to the conclusion that he was the murderer.
When Michael returned to find his home surrounded by police cars, his immediate concern was his son, not his wife, and when informed of her death, he showed little emotion. As is often the case, this only reaffirmed the detectives’ suspicion. Michael then decided to stay in the house that night with his son, which many perceived as somewhat callous given what had taken place there earlier in the day. Michael also eventually resumed sleeping in the very room and on the very bed where his wife had been murdered—another indication to the police and prosecutors of the mindset of a remorseless killer.
Michael adamantly denied killing his wife and fully cooperated with the investigation. He willingly submitted samples of his hair, saliva, and blood, and agreed to take two lie-detectors tests, both of which he passed. Aside from the note and whatever conclusions could be drawn from his behavior and demeanor, not a shred of evidence tied Michael to his wife’s death, despite a hoard of physical evidence recovered from the scene. None of it pointed to Michael. Fingerprints lifted from the doorframe of the family’s sliding glass door failed to match anyone in the Morton family. The source of another fifteen fingerprints found inside the house and near the victim’s body also could not be matched to anyone in the family. A fresh footprint was found inside the Mortons’ fenced-in backyard. And nearby, next to a wooded vacant lot, a blue bandana stained in blood was discovered. None of this evidence could be connected to Michael.
The police not only ignored the striking fact that the forensic evidence pointed away from Michael; they helped bolster the prosecution’s case by turning Elizabeth Gee against him. Leading up to the trial, sheriff’s deputies frequently visited Gee, feeding her information and theories all predicated on Michael’s guilt. At trial, Gee then provided insights into the Mortons’ unhappy marriage, recalling specific fights and instances in which Michael had said unkind things to Christine. Gee also testified that Michael’s stoic demeanor following the murder and the fact that he had cut down his wife’s marigolds two days before her funeral seemed insensitive and suspicious to her.
Michael’s perceived insensitivity helped bolster the prosecution’s case by supporting its contention that he was a depraved, porn-obsessed, sexual deviant who did not care about his wife. According to the state’s theory, after being denied birthday sex, Michael continued watching the adult video, eventually becoming so turned on that he became absolutely incensed by his wife’s refusal to sleep with him. Michael’s sexual frustration and fury came to a head, the prosecution charged, leading him to bludgeon his wife to death. But here’s the real kicker. The prosecution claimed that immediately after murdering his wife, Michael performed one last depraved act: he masturbated on top of her corpse.
Over the objections of the defense, the prosecution showed the sexually explicit video that Michael had rented that night to the jurors. After a six-day trial, the jury deliberated for only two hours before returning a guilty verdict.
Years later, during postconviction proceedings, Michael’s attorneys obtained police reports detailing the investigation of Christine’s murder. The reports revealed the existence of exculpatory evidence that had been withheld from the defense, uninvestigated and ignored leads that should have looked promising, and numerous witness accounts clearly pointing to a different perpetrator.
For example, the morning after the murder, a couple that lived down the street from the Mortons flagged down a sheriff’s deputy to report seeing a man park a green van next to the wooded vacant lot that abutted the Mortons’ backward. This was near the area where the bloody bandana had been found. Another family in the neighborhood contacted the police reporting a similar sighting, but no one ever returned the call. Another witness reported seeing the mysterious man walking through the woods near the van toward the Mortons’ backyard. The police paid very little attention to these reports. In recalling a conversation that had taken place just a couple days after the murder, one witness recalled how the police seemed certain that they had already caught their man, “He didn’t overtly say, ‘We know who did it,’ but he implied that this was not a random event. I can’t remember his exact words, but the suggestion was that the husband did it.”
In the days following the murder, the police also received several telephone calls indicating that Christine’s purse had been stolen. A relative reported that a check he’d written to Christine had been cashed after her death with a signature that appeared forged. Another report indicated that Christine’s credit card had been used in San Antonio two days after her murder. However, because the police were confident that Michael had staged the burglary, neither lead was investigated. Their reasoning is made clear in a note that accompanied the reports: “They seem to think that Chris’ purse was stolen; of course, we know better than that.”
The reports further revealed that at the time of the investigation, the Mortons’ two-year-old son, who by then was in the care of his maternal grandmother, was questioned by the police. Due to the child’s knowledge of the circumstances surrounding his mother’s death, it became abundantly clear that he had witnessed the crime. When asked who was present, the two-year-old accounted only for himself, his mother and a “monster.” When asked whether his father was present, the child very clearly stated no. This was not turned over to the defense. Rather, it was buried.
Twenty-five years after being convicted, Morton was exonerated, in part due to DNA testing. DNA testing on the blue bandana found near the scene of the murder revealed the DNA profiles of Christine Morton and Mark Norwood, a known offender whose DNA profile was available in the FBI’s CODIS database. Norwood was tried and convicted of Christine Morton’s murder in 2013. He was sentenced to life in prison.16
On October 4, 2011, Michael Morton was released from prison. He was officially exonerated two months later and has since been awarded over $1 million in compensation from the state of Texas. Michael later wrote the book Getting Life: An Innocent Man’s 25-Year Journey from Prison to Peace, which detailed how he had been yet another victim of police tunnel vision.
• • •
In 1982, Marvin Anderson of Virginia was another such victim.17 He was identified from a photo lineup as the perpetrator of a rape and robbery by the victim of the crime. Anderson’s photo stuck out like a sore thumb in the array. It was the only photo in color, and the only one with the suspect’s social security number printed across it. Twenty minutes later, with Anderson’s face now firmly imprinted into her memory, the victim selected Anderson from a live, in-person lineup. The identification was the crux of the state’s case at trial, the most powerful evidence against him. Anderson was convicted and sentenced to 210 years in prison.
Anderson had become a suspect through pure happenstance. Among the things the victim recalled about her attacker was that he had mentioned having a white girlfriend. She described the perpetrator as black, and a mixed-race relationship was apparently not that common in Hanover, Virginia, in 1982. The police happened to know one black man who resided with a white woman, Marvin Anderson. Thus, Anderson became the primary suspect.
From that point on, and based on nothing more, the police went after Anderson confidently and aggressively, as if they already had their man. But aside from being a person of color, Anderson did not fit the victim’s description of the perpetrator. Anderson was taller, had a darker complexion, lacked a mustache, and did not have any scratches on his face as described by the victim. Plus, Anderson had four alibi witnesses accounting for his whereabouts during the time of the attack. And he had no criminal record or history of run-ins with the police. But he was nevertheless convicted based on the eyewitness identification by the victim.
The full extent of the tunnel vision that led to Anderson’s conviction was only made clear after DNA testing identified the true perpetrator of the crime twenty years later. The DNA test results not only proved Anderson innocent, but revealed the identity of that perpetrator: John Otis Lincoln. His involvement in the rape had been rumored throughout the community since the time of Anderson’s trial. During the initial investigation, several witnesses reportedly saw Lincoln riding a bicycle near where the crime had occurred. The witnesses also stated that Lincoln had been making lewd and threatening comments toward them that day. The bicycle that Lincoln was reportedly seen riding was later identified by its owner, who verified that it had been stolen by Lincoln on the day of the rape. The only people totally oblivious to Lincoln’s possible involvement were the police, who quickly dismissed him as a suspect early on in their investigation, even though he matched the victim’s description of the perpetrator, was present in the area that day with a stolen bike, and was awaiting trial for another sexual assault at the time. Consequently, Marvin Anderson spent twenty years in prison for a crime he did not commit.
• • •
The Norfolk Four case from Virginia highlights police tunnel vision in the extreme.18 In this case, the police ended up charging no less than seven innocent men with the murder and rape of Michelle Moore-Bosko, before finally charging and convicting the actual culprit. Four of those innocent men served significant prison time, while three had their charges dismissed. In a series of events reminiscent of a Three Stooges episode, the lead detective focused first on Danial Williams and his roommate, Joe Dick, who lived near the victim, because a neighbor said that Williams had been “obsessed” with the victim. The detective, who was known for his aggressive interrogation techniques and was later imprisoned for misconduct in a different matter, coerced both Williams and Dick into falsely confessing, although their confessions did not match one another and flatly contradicted the evidence from the crime scene, including the way the victim was killed.
When the DNA testing from the crime scene excluded both Williams and Dick, instead of questioning his theory of the case, the lead detective went back to them and told them he knew they must have committed the crime with someone else. Which would explain why the DNA did not match them. Under coercion that included threats of the death penalty, Dick ended up offering a new name, Eric Wilson. The detective then arrested Wilson and coerced him into a false confession as well. But when the DNA from the crime scene didn’t match him either, the detective went back to the well again and told the men they had to name who else was with them. Under great pressure, Dick eventually named Derek Tice, and Tice was also promptly arrested, and he too falsely confessed. When Tice’s DNA did not match the crime scene evidence either, he was pressured to name more men, and so he provided the names of the three individuals whose charges were later dismissed. (They had not falsely confessed.)
Although the evidence from the crime scene strongly suggested that only one man committed the murder and rape, the lead detective’s theory of the case changed as each suspect was excluded by DNA testing and a new man was offered up as an accomplice under coercion. Eventually, the detective’s theory expanded to include eight different men participating in a gang rape and murder—a theory that was wholly contradicted by any unbiased analysis of the crime scene, including the inconvenient fact that only one man’s DNA was present. Yet the state went forward and obtained convictions against Williams, Dick, Wilson, and Tice, based primarily on their confessions. That DNA from the crime scene came back to match Omar Ballard, who had no connection to the other men and who confessed to having committed the crime alone. The Norfolk Four were eventually released and fully pardoned after great public pressure was brought to bear on their case.19
• • •
In April 1989, a female jogger was brutally raped and nearly beaten to death in New York City’s Central Park. An hour later, before the unconscious jogger was discovered, the police chased down a group of thirty teenage boys who were reportedly harassing people in the park in a moblike fashion. Five of them were apprehended and charged not only with participation in the mob, but also the jogger’s rape. The boys, Antron McCray, Kevin Richardson, Yusef Salaam, Raymond Santana, and Korey Wise, ranged in age from fourteen to sixteen. Four of them subsequently confessed, which led to the convictions of all five. However, several years later, DNA testing would prove that all five of them were innocent and that a man named Matias Reyes, a serial rapist acting alone, was actually responsible for the crime.20
The jogger’s attack received a great deal of media attention due to the nature of the case. It involved a sexual assault on a white female victim who was a wealthy, Ivy League–educated investment banker, and a gang of black and Latino teenage perpetrators. The case even caught the attention of none other than Donald Trump—who famously called for the immediate execution of all five boys. The case was as high-profile as they come, and the higher a case’s profile, the more pressing the need to solve it. Under intense external and internal pressure to obtain convictions, those handling the jogger case became hell-bent on securing confessions from the boys.21
No one witnessed the attack on the jogger, and because she had sustained a head injury during the attack that caused her to black out, the jogger could not provide a description of the assailant. Aside from the boys’ presence in the park and their alleged participation in the mob, the police had no reason to suspect them of the jogger’s rape. Indeed, there was nothing indicating that the jogger had been attacked by more than one assailant. However, the police subjected the boys to intense and lengthy interrogations, ranging from fourteen to thirty hours.22
Their four confessions were inconsistent not only with one another, but with demonstrable facts of how the attack occurred, which should have led the Manhattan district attorney to question their reliability. But instead they were the primary evidence used to convict the boys. The confessions were admitted despite defense arguments that the use of overly manipulative and coercive interrogation tactics, including threats, physical abuse, and promises of release, effectively rendered them involuntary. Conveniently, only the confessions were recorded and not the actual interrogations preceding them, meaning there was no way to prove the extent of the coercion.
Aside from the confessions, there was no evidence linking the five boys to the jogger’s rape.23 In fact, the physical evidence suggested the contrary—that the boys were innocent. At the time of their trials, an early form of DNA testing was performed on several samples from the crime scene. The Manhattan DA’s office reported to the press and the public that the results of the testing were inconclusive, but, in truth, the testing had failed to find a match between the defendants and the DNA samples: all five defendants were excluded. Furthermore, the testing showed that semen obtained from the victim’s cervix matched semen found on her sock, which tended to suggest that there was only one perpetrator. Yet, rather than cease prosecution of the boys, the DA, in effect, suppressed the results by construing them as inconclusive.24
At trial, each of the five boys was convicted; they received five to fifteen years apiece. In 2002, Matias Reyes, who was serving a life sentence in a New York State prison, confessed to the rape and assault of the Central Park Jogger. Reyes had been regarded as one of New York’s most notorious serial rapists. New DNA testing with advanced technology confirmed that the semen on the jogger’s sock belonged to Reyes.25 On December 19, 2002, the convictions of Antron McCray, Kevin Richardson, Yusef Salaam, Raymond Santana, and Korey Wise were overturned. Wise, the only one of the five boys to be charged as an adult, served eleven and a half years in prison. The Central Park Five received more than $40 million in compensation from New York City.
• • •
In the prosecutor’s office where I worked, there was no general awareness of the dangers of tunnel vision. Building a case and making each piece of new evidence fit our preexisting theory was often a group activity. We had fun picking each other’s brains and brainstorming on how to spin evidence to fit our original hypothesis. It was a game. There was no sense of objectivity about the process whatsoever. There was never an attempt to disprove our original hypothesis. Rather, the point of the game was to see how clever you could be in making all new evidence fit your original hypothesis, thus making the case stronger. To the extent we played devil’s advocate and put forth contrary arguments, it was only so we would be prepared to shoot down such arguments in court.
Not only are there no systematic policies in place at most police departments and prosecutor’s offices to combat tunnel vision, but police are actually trained to engage in tunnel vision. The famous Reid interrogation technique, for example, which most police officers in this country are trained in, encourages police officers to assume the guilt of those under interrogation and to confidently inform the suspect that the officers know they are guilty. When suspects under interrogation try to offer up innocent explanations or alternative theories suggesting their innocence, officers are instructed not to listen and to gruffly cut off all such attempts. As one scholar has noted: “The very notion of the Reid ‘interrogation,’ therefore, expressly embraces the foundation problem with tunnel vision—a premature conclusion of guilt, and an unwillingness to consider alternatives. In this context, however, the tunnel vision is not inadvertent, but deliberate; police are taught that this is the way to advance their investigation. Cognitive biases are openly encouraged.”26
• • •
Those like myself who work on postconviction innocence cases see another type of tunnel vision in our criminal justice system—defense attorney tunnel vision. In case after case, we see that the defense attorney at our client’s original trial did very little work and often performed no independent investigation into the facts. They often simply just accepted the foundation of the state’s case and worked within it, merely cross-examining the state’s witnesses as their primary form of defense, without performing any of the independent investigation necessary to offer evidence supporting an alternative hypothesis. Part of the reason for this is that defense attorneys typically don’t have the resources at their disposal to challenge the prosecution’s theory of the case. But it often goes beyond that. As one postconviction innocence lawyer has noted:
We’ve all read about tunnel vision and how it affects police investigations and prosecutors’ decisions. Defense lawyers also fall prey to cognitive biases. They often see their clients as the jurors and judges saw them—guilty. And they often see their clients’ cases as unwinnable. Because of these cognitive biases we often fail to raise the claims that could eventually lead to exoneration. Lawyers are wise to investigate all the facts and test all the theories, withholding judgment until the investigation is complete. Only when every theory has been debunked and every favorable relevant fact turns out to be false, can a post-conviction lawyer throw in the towel.27
Another scholar has noted, “After a year toiling as public defenders, most stop believing that any of the people they represent are actually innocent. In fact, in most criminal courtrooms, most people present believe that the police got it right and that the trial will simply confirm the defendant’s guilt.”28
This phenomenon was seen starkly in the Netflix docu-series Making a Murderer. After Brendan Dassey gave police what appears to be a demonstrably unreliable confession that he was Steven Avery’s accomplice in the murder of Teresa Halbach, he was arrested. As the film progresses, the viewer sees that Dassey may very well have been entirely innocent, even though he was eventually convicted, and that he had many avenues available to make a strong defense. But his first court-appointed attorney, who represented Brendan at his initial appearance before bowing out due to a conflict of interest, stated in open court that Brendan had participated in the crime before even speaking at length to his client or doing any investigation. He just blindly accepted the prosecution’s theory of the case. Dassey’s next attorney, Len Kachinsky, did the same thing. During Kachinsky’s representation of Dassey, Kachinsky never investigated the case nor made any effort to listen to what Dassey was telling him—that he was innocent and had falsely confessed. In fact, when Dassey protested his innocence, Kachinsky sent his defense investigator into the jail to intensely pressure Dassey to give up his claims of innocence. The investigator eventually got Dassey to sign another confession that mirrored his first one.
Part of the problem is that public defenders and court-appointed defense attorneys are conditioned to simply accept the state’s theory because they do not have the funds to hire investigators, or expert witnesses, to perform their own full-scale investigations. But another part of the problem is defense attorney tunnel vision.
In my work with the Ohio Innocence Project, I cannot count the number of times when, after we have opened a case and begun reinvestigating it, I have called the client’s original trial attorney to seek information about the case, only to receive a condescending laugh and a comment like, “That guy? The Ohio Innocence Project is looking into that case? That guy is totally guilty.” When I have then asked them how they had investigated the case, if at all, and whether they had looked into the promising angles that the OIP had recently uncovered, the answer is often, “No. I didn’t do any of that because it just seemed clear that the guy was guilty. I’m not gonna ask the judge for funds for investigators and stuff like that in a case where he’s so obviously guilty. I’d get laughed out of court.”
In cases when we are eventually able to build powerful evidence of innocence, the reaction from defense attorneys is mixed. Some are concerned, or embarrassed, and offer to help in any way possible. But many continue to believe that their former client is guilty. They cannot accept the fact that they might not have done enough, that their lapses may have caused an innocent person to be convicted. Tunnel vision, it seems, can be hard to shake, even for a lawyer charged with protecting the rights of the victims of tunnel vision.
In any field there are, of course, bad apples. In the criminal justice system, no category of actor is immune to this problem. The public has been shocked in recent years when videos have surfaced of cops shooting and killing unarmed suspects who weren’t fighting back.29 There have been publicized cases of police officers framing innocent people, planting evidence, or even torturing suspects with electric shocks to obtain confessions.30
There has been no shortage of stories about forensic lab technicians who have been caught routinely fabricating results—rendering conclusions favorable to the prosecution without even running the tests. Massachusetts forensic lab technician Annie Dookin, for example, was sent to prison in 2013 after it was discovered that she routinely faked test results in thousands of cases.31
Defense lawyers have gone to prison for bribing witnesses or encouraging them to leave town so they couldn’t testify against the lawyer’s client at trial. Other defense lawyers have become immersed in the criminal activity of their clients, apparently unable to resist the monetary temptations.32
Even judges are sometimes bad apples. One judge in Pennsylvania, in what’s known as the “Kids for Cash” scandal, was recently convicted for taking payoffs from a private prison company for each juvenile he sentenced to a juvenile detention center, thereby enriching the company by giving them yet another juvenile inmate it could charge the state for incarcerating.33
The public understands that sometimes bad people—people with personality disorders or criminal minds—infiltrate the criminal justice system as police officers, prosecutors, defense lawyers, judges, or in other roles. Just as disturbed people sometimes become doctors, accountants, or teachers. It can’t be completely avoided, despite our best efforts to keep them out. We all know that.
But most of the injustice in the criminal justice system results from the actions of good people, not the rare bad apples. Indeed, good police and prosecutors can cause tragic injustices when they suffer from tunnel vision. A certainty that they have identified the guilty party can sometimes lead police or prosecutors to cut corners—or even engage in acts of outright dishonesty—to obtain the result that they know is the right result in the big picture. This phenomenon is called “noble-cause corruption.”34 Police and prosecutors who engage in noble-cause corruption are motivated by a sense of their important mission to protect the public. It is an “ends-based investigative culture that prompts investigators to blind themselves to their own inappropriate conduct, and to perceive that conduct as legitimate in the belief that they are pursuing an important public interest.”35
Noble-cause corruption is different than outright corruption—such as accepting bribes to dismiss charges—because those who engage in it are not seeking a personal advantage, like money, but believe they are doing it for a laudable cause. A cop may think, “It would be a tragedy to the victims and society at large if this guilty murderer gets away with it. I need to bend a few rules to bring justice to the victims and to save the life of his next would-be victim.”
Defense attorney Dean Strang, who in Making a Murderer raised the defense that the police had planted evidence against his client Steven Avery, captured this understandable human temptation: “I didn’t see [the cops] plant evidence with my own two eyes. . . . But do I have any difficulty understanding what human emotions might have driven police officers to want to augment or confirm their beliefs that [Avery] must have killed Teresa Halbach? I don’t have any difficulty understanding those human emotions at all.” Strang also commented, “It is really unusual—and an abandoned act of evil—when a police officer sets out to frame someone who he knows to be innocent.” Strang sees any police misconduct in Avery’s case as arising not from bad intentions, but from a determination to convict the right person.36
Most everyone has been asked at one time or another, “What is the first thing you would do if you had a time machine?” I have heard people answer, “I would go back and stop Hitler before his rise to power, even if it meant killing him.” I can understand the temptation and human desire to want to do this. Indeed, in a recent poll, when asked whether they would kill a baby Hitler if they had a time machine, 42 percent said yes, 30 percent said no, and 28 percent said unsure.37
And so it is with police officers and prosecutors who engage in noble-cause corruption. If one is convinced that a suspect under investigation is a heinous murderer who may get off scot-free unless a few rules are broken, the temptation to break those rules can be overpowering. It’s human. In Robin Hood–type fashion (“steal from the rich and give to the poor”), law enforcement officers can convince themselves that their minor acts of dishonesty are necessary to promote an important public good, and that if they don’t engage in such acts the public will suffer greatly.
But while police and prosecutors tend to always believe they are pursuing a heinous criminal, and that their suspect is dangerous and evil, history has proven that they are often wrong.
There have been many stark examples of noble-cause corruption, such as planting evidence or hiding evidence that is favorable to the defendant. Suppressing favorable evidence—what we legally called Brady violations—is so common that a judge on the U.S. Court of Appeals for the Ninth Circuit recently wrote in a judicial opinion that it has become an epidemic: “I wish I could say that the prosecutor’s unprofessionalism here is the exception, that his propensity for shortcuts and indifference to his ethical and legal responsibilities is a rare blemish and source of embarrassment to an otherwise diligent and scrupulous corps of attorneys staffing prosecutors’ offices around the country. But it wouldn’t be true. Brady violations have reached epidemic proportions in recent years, and the federal and state reporters bear testament to this unsettling trend.”38 Several of my innocence cases have involved egregious Brady violations, including the Gillispie case discussed earlier. My Ohio Innocence Project clients Wheatt, Glover, and Johnson were also victimized by atrocious Brady violations.
Wheatt, Glover, and Johnson were all convicted of murder based on the identification of a single eyewitness and alleged gunshot residue found in a couple small areas on their clothing shortly after a shooting occurred on a street in East Cleveland. Gunshot residue is what is often deposited on a person’s hands or clothing from the “blowback effect” after he or she fires a gun. The prosecution had offered no motive for why the three men would have shot and killed the victim, and no other evidence. Calling themselves the “East Cleveland 3,” the defendants adamantly protested their innocence both before and after their convictions.
Several years into their prison sentences, the eyewitness recanted, saying that she had been pressured by the police into making the identification. The judge who tried the case originally, and who later admitted that she had been concerned about the case since the jury handed down the three men’s convictions years earlier, heard this witness’s recantation, found it credible, and overturned the trio’s convictions. The prosecution appealed, however, and the appellate court reinstated their convictions, finding that the gunshot residue evidence was still sufficient to implicate the East Cleveland 3 in the murder.
At that point, the OIP got involved. We got involved because we knew that the gunshot residue evidence—the sole evidence maintaining their convictions—was highly problematic. In the years since the trio’s convictions, it had been discovered that the test used at the time of their trial was prone to produce “false positives”—meaning that the test would signal the presence of gunshot residue when that residue was in fact some other material found in the environment having nothing to do with firing a gun.
More important, the East Cleveland 3 had had their hands and clothing swabbed for gunshot residue after they had been handcuffed and taken to the police station in squad cars and interrogated. Recent studies revealed that police cars and police stations—even police uniforms—are awash in gunshot residue due to the amount of firearm training police officers must undertake on an ongoing basis.39 This causes contamination, as suspects who are transported in a squad car can get gunshot residue on their clothing merely from being in the car. As one commentator has noted, if a suspect is arrested and taken to the police station in a squad car, and then swabbed for gunshot residue at the station, a subsequent test finding a particle of gunshot residue indicates guilt no more than a particle of sand on the bottom of a suspect’s shoe indicates he had just been on a beach.40 As a result of these findings, law enforcement across the nation have been advised to change the protocol for sampling such residue, and to take swabs from suspects when they are first apprehended and before being placed in a squad car, or only after bagging their hands to protect them from contamination until swabbing can be performed at the station.41
Based on the eyewitness’s recantation, and these new scientific understandings undermining the probative value of the gunshot residue evidence, we moved to overturn the East Cleveland 3’s convictions.
We lost. We also lost in the court of appeals. The case went cold from our perspective, and the East Cleveland 3 remained in prison.
A few years later, we filed a public records request for the original police file. The East Cleveland 3 had been trying from prison to get the file for years but the police had refused to turn it over. But this time, however, for reasons we don’t know, we unexpectedly received the file in the mail. And what we found was alarming. In the file were police reports made during the initial investigation that revealed two things. First, a group of unknown males had been trying to kill the victim and his brother in the days leading up to the murder. In fact, the victim had been shot at just a few days before in drive-by fashion, similar to how he was ultimately killed. The car driven by the shooters in that incident was different from any car associated with the East Cleveland 3. And during the original investigation, the police showed photos of the East Cleveland 3 to the victim’s brother, and he did not identify any of them as the drive-by shooters in the previous attempt. So it appeared that someone else had a motive to kill the victim at the time he was murdered. And he wasn’t any of our clients.
Moreover, while the eyewitness in the original investigation (who later recanted, saying she had been pressured by police) had identified one of the East Cleveland 3 as the shooter, two other eyewitnesses had given statements that completely contradicted that witness, and, in fact, had identified the shooter as someone else entirely—a man those witnesses knew from the neighborhood.
All of this evidence pointed decisively to the East Cleveland 3’s innocence. But it had been hidden by the police and, in violation of their constitutional rights, never turned over to the defense prior to trial. The defense attorneys at the original trial could not interview these witnesses or call them as witnesses, because they simply didn’t know about them.
A few years after their convictions, and while they were in prison, the East Cleveland 3 sent a public records request to the police department asking for the full police file. Their request was ignored. But when we finally got the police file years later, it contained a letter from the prosecutor’s office to the police department, sent after the East Cleveland 3 had asked for the file, telling the police department to ignore the request. In fact, the letter demanded that the police file be sent to the prosecutor’s office for safekeeping. However, for reasons we don’t know, the police department did not send the file to the prosecutor’s office as the letter demanded. That is how we were able to obtain it years later when we filed our own public records request with the police department.
Thus, someone in the police department, before the original trial, had made the decision not to disclose the exculpatory police reports to the defense, though the Brady rule required them to do so. And then the prosecutors made the decision after trial to ensure that the police file would never see the light of day.
Talk about covering your ass.
It’s perhaps easy in hindsight to cast these actions as attempts to manufacture and maintain a bogus conviction. Acts of evil, even. But I believe that the most likely reason for them is that those who committed these acts suffered from severe tunnel vision, were convinced that the East Cleveland 3 were guilty, and felt that they were doing a public service to bend some rules to guarantee that the East Cleveland 3 would be convicted and stay in prison for life. That’s noble-cause corruption.
Armed with new evidence of these Brady violations by the prosecutors and police, we once again sought to exonerate the East Cleveland 3. In March 2015, the trial judge threw out their convictions and the trio were finally freed after they’d each spent twenty years in prison. The decision was upheld on appeal, and later the prosecution dropped all charges against them.
• • •
Noble-cause corruption occurs in more subtle forms as well. I have previously described the problem of “synthesized testimony.” This occurs when police or prosecutors take witnesses who initially provided details helpful to the suspect and, convinced because of tunnel vision that the witness must be wrong, ask pointed questions of the witness suggesting that his or her memory is off. Some witnesses take the cue and begin to question their memory. Sometimes they will change their statements to correspond with what it is clear the police believe must have happened. Then they “remember” it this way.
But even worse is when these initial comments helpful to the defense are not written down by law enforcement. The Brady rule requires law enforcement to disclose to the defense all materially helpful evidence uncovered during an investigation. But if a witness gives an initial statement helpful to the defense, and then changes it to fit the police’s theory after pointed questions are directed at him or her, the initial statements are often not written down. At trial, when the witness testifies, the defense lawyer will have no idea that the witness initially provided the client with an alibi, but then modified his or her statement only in response to police questioning.
As I said before, as a new prosecutor I was taught by mentors not to write down parts of a witnesses statements that I believed were incorrect (and harmed my case) until after the witness “came to Jesus” and told the “correct” story. I was told that if the initial “incorrect” statements were written down, I’d have to turn them over to the defense, and the defense attorney would then be able to tear the witness to shreds on the stand for changing her story. And that would be a shame, I was told. A guilty criminal shouldn’t get off because of some careless witness who opened his mouth and blurted something out before thinking it through. The practice was described to me as a way of ensuring that the guilty got what they deserved, and that they didn’t get off on some stupid mistake that a defense attorney could unfairly exploit at trial by blowing it out of proportion.
Noble-cause corruption also occurs in even subtler and more intangible forms. As depicted in Making a Murderer, the day after sixteen-year-old Brendan Dassey gave a confession implicating himself as the accomplice of Steven Avery in the gruesome murder of Teresa Halbach, prosecutor Ken Kratz held a press conference in which he gave dramatic, chilling details of the crime as allegedly revealed by Dassey’s confession. Kratz told television viewers that the details he was about to provide were so gruesome that anyone under age fifteen should change the channel. He then took viewers through his theory of how the murder occurred, based on Dassey’s confession, and seemed to revel in the drama of the gory and sordid tale.
But—wait—Dassey’s “confession” was recorded, and the videotape suggested that Dassey knew nothing about the crime. Rather, it appeared that he had just given the interrogators what they wanted to hear in response to leading questions. Indeed, Dassey’s confession appears on the videotape as the prototypical false confession. Dassey would give wrong answers to questions until the interrogators fed him the “right” answer, which he would then adopt. For example, the interrogators, wanting Dassey to reveal the crucial fact not yet known to the public—how Halbach was killed—asked Dassey, “What did you guys do to her head?” Apparently not knowing the right answer, Dassey could not provide one. After the officers repeated the question several times, Dassey finally guessed, “We cut her hair.” But because that was not the answer the police wanted, they then asked him repeatedly, “What else did you do to her head? . . . Come on Brendan, what else do you do to her head?” Eventually, Dassey guessed again, “We punched it.” Because this was still not the right answer, one of the officers ultimately lost his patience: “Okay, I’m just gonna come right out and say it. Which one of you shot her in the head?” At which point Dassey answered, “He did,” meaning Steven Avery.
The videotape also makes clear that Dassey did not understand the importance of what he was doing or saying, as he asked the officers at the end of the interrogation whether he would be able to get back to school by 1:30 because he had a project due. Of course, he was not allowed to leave and he has remained in prison since that date. After the interrogation ended, and the police officers stepped out of the interrogation room, Dassey’s mother entered the room and asked her son why he had confessed to something he didn’t do. Dassey answered, “They got in my head.” A subsequent search of the Avery property failed to corroborate the wild story that Dassey told, or more accurately, the “synthesized” story he adopted in response to the police questioning. Dassey later told his mother that he had merely “guessed” at what answers the police wanted during his interrogation.
But despite the obviously problematic nature of this confession, Ken Kratz went on television and told a gruesome tale to the public based on it. This had the effect of sealing Steven Avery’s fate in the eyes of the public from which his jurors would eventually be taken. Making a Murderer cut from Kratz’s press conference to news interviews with various members of the public. Some of these people said that while they had not been sure before the press conference whether or not Avery was guilty, after hearing about the “confession,” they now firmly believed he did it. Though Dassey’s confession wasn’t even admitted into evidence at Avery’s subsequent trial, Kratz had ensured that it was firmly entrenched in the public’s consciousness, and thus in the minds of Avery’s jurors, before even checking its reliability or knowing that it would be admitted into evidence.
Kratz’s tunnel vision was so strong that he ignored the facial unreliability of the confession at his press conference. And I’m sure that, in his mind, there was no chance that the confession was untrue. Kratz seemed supremely confident in what he was saying, and by sharing the gory details he made the press conference that much more tantalizing and memorable to the public. Experts have called Kratz’s conduct unethical because it contaminated the local jury pool. But no bar proceedings were instituted against him for violating the pretrial publicity rule.42 Most prosecutors who engage in this conduct are not disciplined. Indeed, in the counties surrounding the city where I live, prosecutors routinely engage in this same type of pretrial publicity stunt when they bring charges in high-profile cases, which, of course, contaminates the jury pool in their favor.
When prosecutors do this sort of thing, however, they always think it’s for the good. They believe they are fighting for a righteous cause to protect the public and to ensure that an evil person is punished. And holding a press conference of this nature is a good way to move one step closer to “justice.”
• • •
There are probably some cases in which, if the officer didn’t bend the rules a little, the guilty person might end up getting off scot-free. And I understand that if a person’s close family member was murdered, and he or she felt certain that a particular suspect was guilty, he or she might want the rules bent a little to ensure the perpetrator was brought to justice. That temptation is natural and understandable.
But our criminal justice system shouldn’t be grounded in the basest and most retaliatory instincts of the victims at their time of deepest shock and grief. If we based our system of justice on such instincts, then we might as well shoot and kill those arrested for heinous crimes as soon as they are apprehended. Although my mind would tell me different, my basest instincts would want that done to anyone who I believed had just murdered one of my children. And I’m sure many parents would feel the same way. We cannot, of course, follow such retaliatory instincts, but must allow our deeper sensibilities and more rational understandings of fair process to prevail.
Turning a blind eye to noble-cause corruption would also mean that police officers and prosecutors, rather than the judicial system and the trial process, would be the ones who seal the fate of those charged with crimes. And when a police officer or prosecutor engages in an act of noble-cause corruption that seals the fate of a suspect—such as, for example, by manipulating a witness’s memory to invalidate the suspect’s alibi—it happens behind closed doors rather than in public view as when a person is convicted during a trial. As a result, the conduct cannot be reviewed or scrutinized. It never sees the light of day.
But most important, police officers and prosecutors are sometimes wrong, even when they’re sure they’re right. Because they are human, they are prone to suffering from tunnel vision, and this can cause them to focus on innocent suspects. If we truly believe the adage that “it’s better that ten guilty persons escape than that one innocent suffer,” upon which our criminal justice system is supposedly based,43 then this type of behavior simply cannot be pardoned.
The intense pressure on law enforcement to solve crimes, which often brings with it tunnel vision and noble-cause corruption, has through time created entire categories of evidence that are systemically corrupted. Examples include eyewitness identification testimony, confession testimony, incentivized-witness testimony, and forensic testimony. When I say that each of these categories of evidence is corrupted, I do not, of course, mean that every single eyewitness identification introduced in a courtroom is inaccurate, that every confession introduced is false, or that every statement by a forensic “expert” or incentivized witness is misleading or fabricated. Far from it.
But these categories of evidence are: (1) sometimes unreliable or false; (2) always presented at trial by the prosecution as solid and entirely reliable in the case at hand; and (3) capable of being substantially improved and made more reliable. Yet we fail as a system to take steps to improve them. If certain categories of evidence were sometimes unreliable and we were honest about that fact to jurors, and there was nothing more we could do to improve them, that would be OK in my book. You can’t just stop prosecuting people because some forms of evidence aren’t perfect. And in such circumstances, we would be doing the best we could do and being honest about our limitations. But instead we overstate the reliability of these forms of evidence, which, in essence, amounts to hoodwinking the jurors, while simultaneously failing to take sufficient steps to improve their reliability.
Studies show that simple steps could be taken to improve the reliability of eyewitness identifications.44 Eyewitness errors occur for a variety of reasons. Error can enter the process when an officer, convinced she knows who the perpetrator is and desperately wanting the eyewitness to select him, gives overt or slight hints to the witness as to which photo she should select. This can be very subtle, such as, “I see you are looking closely at #2. What about #2 makes you think it could be him?” Or it could be unconscious, such as when each time the witness’s eyes settle on photo #2, the officer suddenly tenses up, holds his breath, and leans over the suspect’s shoulder to get a better look in anticipation of the “right” choice being selected, unaware that he is sending a strong signal to the witness that #2 is, indeed, the “right” answer. We speak more with our bodies than we do with our mouths. Professional poker players wear sunglasses and hats for this very reason—even those with practiced “poker faces” can’t help but inadvertently tip their hand.45
Or the officer might tell the eyewitness after she makes a choice that she selected the suspect. As we’ve seen, after an eyewitness selects a photo, confirmatory feedback of this nature has the effect of increasing a witness’s confidence level. This can easily turn what would have been “I picked #2 because he looks the most like the perpetrator, but I can’t be sure,” into “I’m 100 percent positive it is #2.”
To curb such alterations of memory by police, the identification process must be double-blinded. This means that the officer administering the photo lineup shouldn’t know which of the six photographs depicts the police’s suspect. An officer not assigned to the case should thus be the one who performs the lineup process with the eyewitnesses. There are also easy methods that have been devised that allow the officer assigned to the investigation to administer the lineup but blind him to seeing the photographs, thereby reducing his ability to subtly sway the outcome.
The double-blind process is elemental to the scientific method, because it eliminates bias.46 If a test is not performed blindly, we simply don’t accept the results in other fields. To take a ridiculous example, if a Pepsi vs. Coke taste test is performed on customers strolling through a mall, and the person administering the taste test works for Coke and knows which cup contains Coke, we would call the results unscientific and flatly reject them. Even people doing taste tests at malls know that to be reliable a test has to be double-blinded. Yet we convict our own citizens of serious crimes, and sometimes even send them to death row, based on a multiple-choice test—which of the six photographs is the perpetrator?—that doesn’t follow a core principle of the scientific method designed to remove subtle bias and maximize reliability.
In addition, there must be a process in place for immediately memorializing the eyewitness’s confidence level as soon as she says, “I’m picking #2.” Witnesses who are 60 percent confident at the time they make their pick, for example, are often 100 percent confident in their selection by the time they testify at trial, due to postselection confirmatory feedback. The most obvious confirmatory feedback occurs when the person the eyewitness selected is subsequently arrested and the eyewitness is asked to testify against him at trial. In the meantime, she’s likely seen the suspect’s face on the news, or learned that the prosecution has substantial other evidence against him. But if her actual confidence level is recorded by a double-blinded officer immediately after she selects a photograph, this problem is alleviated to a large extent. The defense attorney will be able to point out that the eyewitness was only 60 percent sure when she made the pick, and that the increase in her confidence after her selection was brought about by external factors.
Research shows that eyewitnesses should also be shown photographs one at a time rather than at the same time.47 This is called the “sequential method.” The theory behind this process is that when a witness sees all six photographs together, her mind makes comparative judgments, and she ultimately selects the person in the lineup who most resembles the perpetrator. But the process of selecting a perpetrator from six different faces alters her memory, so that, once she makes a selection and says, “It’s #2,” she then “remembers” the face of #2 as the perpetrator rather than the face of the person who actually committed the crime. And when the police are wrong about their suspect and the actual perpetrator’s photo is not in the lineup, this can cause a witness to select an innocent person and then become convinced that this innocent person is the perpetrator. This problem has been documented not only in clinical studies but in case after case.
However, studies show that when an eyewitness is shown photographs sequentially, her mind is forced to compare only the photograph in front of her to her memory of the perpetrator’s face, reducing the comparative-mixing effect of the traditional six-pack lineup format. This increases the chance that she will say, “The perpetrator is not in this lineup,” in cases where the police have developed suspicions against an innocent person. While many psychologists and researchers have recommended the sequential method, in 2014 the National Academy of Sciences issued a report stating that more research is needed before it would recommend this process.48
The eyewitness identification process can also be made fairer and more reliable at trial by including jury instructions explaining to jurors the inherent problems with memory, and the fact that eyewitnesses are not always correct. And it can be improved by the liberal allowance of memory and psychology experts at trial to educate the jury about the process of identification and the proper weight to be given to the testimony of eyewitnesses. The identification process should be videotaped from beginning to end, to document the conditions associated with the initial identification.49
In its current form, however, eyewitness identification testimony is systemically corrupted because, by and large, these procedures are not followed except in limited jurisdictions. And the flaws inherent in eyewitness testimony are usually not presented to the jury. Rather, prosecutors always tell jurors they can take this type of evidence to the bank, and that the eyewitnesses who testified in this case are 100 percent reliable.
The problem with false confessions can be significantly alleviated rather simply: interrogations must be videotaped from beginning to end. In this way, jurors can see the stages of the interrogation and whether the confession that resulted came about naturally and convincingly, or by the police pressuring the suspect and feeding him information about the crime for him to parrot back. Today, police typically do not videotape an entire interrogation; they turn on the camera only after hours of pressure, when the suspect is ready to recite from beginning to end his “synthesized” confession. The jury sees only the final product, often scripted, and not what went on before, which resulted in the manufactured product.
Branden Dassey’s confession to participating in the murder of Teresa Halbach, and implicating Steven Avery as the mastermind of the killing, was videotaped from beginning to end only because Wisconsin had recently made this a requirement in the “Avery Bill.” This bill came about because of Avery’s wrongful conviction and DNA exoneration in the earlier Penny Beernsten rape case. Based on my experience, I know that if Dassey’s interrogation had not been videotaped from beginning to end, the nuances that demonstrate its unreliability would have never come to public light. The police would have presented it at trial as a solid confession that Dassey gave without prompting and without being fed information about the crime. Because of tunnel vision and noble-cause corruption, police do not see themselves as feeding information to suspects during interrogations. They are always convinced the person they are interrogating is guilty, and when the confession finally materializes they are confident that it’s reliable.
Dassey was convicted at trial even though the unreliability of his confession was, for the most part, depicted to his jury. But in some cases the existence of an interrogation videotape can facilitate an acquittal when that appears to be the proper outcome. And Dassey’s conviction was overturned by a federal court because of the way the police interrogation went down, as depicted on the videotape.50 Dassey wouldn’t have had a snowball’s chance in hell if his interrogation hadn’t been recorded in full.
As with eyewitness identification, the reliability of confession evidence could also be improved through proper jury instructions and the liberal allowance of expert witnesses to discuss the psychological pressure points inherent in the interrogation process.
Departments should also explore alternatives to the Reid technique of interrogation, such as the HIG or the PEACE methods. The Reid technique is prone to false confessions because it creates a highly coercive environment where the interrogator assumes the guilt of the suspect and tries to force words into the suspect’s mouth (whether true or not, with “truth” defined by the interrogator, who is often blinded by tunnel vision). Moreover, the Reid technique tends to cause the guilty—the savvy criminals—to clam up or demand a lawyer, leaving the innocent, who feel they should talk because they have nothing to hide, the more vulnerable to manipulation and false confessions. The HIG and PEACE methods, in contrast, focus on rapport building and getting the suspect to talk freely.51 If done correctly, even savvy criminals, many of whom would “lawyer up” under the Reid technique, can be convinced to talk. The HIG technique (named after the High-Value Detainee Interrogation Group that created it) was developed in recent years by the FBI, CIA, and Pentagon as they searched for a new interrogation method to use in the war against terror that would be more effective than the Reid technique. The similar PEACE method of interrogation (which stands for Planning and preparation, Engage and explain, obtain an Account, Closure, and Evaluation) has been in use in the UK for over a decade. Under these techniques, interrogators go to great lengths to make the suspect feel comfortable and “prime” him to open up and let the information flow. His statements are then compared with readily identifiable objective facts. Many guilty suspects eventually dig their own graves. Research has found these techniques more effective than the Reid technique at obtaining valuable confessions. And more important, they are less prone to false confessions. The LAPD has experimented with the HIG interrogation method, but so far, they are the only domestic police department to have done so.52
But in its current form, confession evidence is systemically corrupted because it’s presented by prosecutors as foolproof, when in reality it is often unreliable or misleading. Simple steps have not been taken in most jurisdictions to improve the accuracy of confession evidence.
Our criminal justice system relies heavily on the use of incentivized witnesses, often called informants or “snitches” in legal slang, who testify for the prosecution in exchange for leniency in their own criminal cases. These witnesses are usually criminals who have a great incentive to lie—to tell the prosecution what it wants to hear—so that they can cut a deal for themselves. Although common sense suggests the testimony of such witnesses should be considered by jurors with a grain of salt, prosecutors always tell the jury that the snitch in this particular case is telling the truth and passionately urge the jury to believe him. With the deference that jurors typically give to prosecutors, who they figure know more about the case than they do, lying incentivized witnesses cause innocent people to be convicted. Indeed, in the cases of the first 325 DNA exonerations, 14 percent of the wrongfully convicted people had been convicted in part by the testimony of such witnesses.53 According to the National Registry of Exonerations, more than 50 percent of wrongful conviction cases nationally have involved perjured testimony against the defendant, a good percentage of which involved incentivized witnesses.54
When I was a prosecutor, I never suffered for the lack of a snitch. In some cases, multiple criminals or co-defendants would compete for the right to testify against the defendant in my upcoming trial. I would interview them—in reality, audition them—to see who would be the best witness in my case. I would ascertain which one could give me the best details and who could come across the most convincingly. The prosecutors in my office had a little routine we went through to teach our snitches how to come across to the jury as truthful. We told them, “Admit all your past crimes. Don’t minimize anything. Don’t argue with the defense attorney—just admit the stuff she throws at you with your head down, ashamed. The point is to convince the jury that you have turned over a new leaf—you are starting a new life. And that you’ve realized that your first step toward an honest life—the way to get your life back on track—is to come clean and tell what you know about this case.”
We had a specific image we shared with our snitches to describe how they were supposed to come across on the stand when telling their story—humble, remorseful for all their past deeds, with head down and “hat in hand.” “Yes, yes,” they were supposed to say, “I’ve done some bad things. But I realized that to start over, I had to start somewhere. And this is my way of joining the good side, of finally going against the bad guys and helping to put them in prison instead of being one myself. It took a lot of courage for me to do this. It isn’t easy.”
Of course, the whole thing was a rehearsed act. A dramatic production of some twelve-step, self-discovery, and self-improvement program that the jury would buy, thus believing the snitch. We prepared these witnesses like theatrical directors did with their actors down the street on Broadway. If pulled off properly, the jurors wanted to believe the snitches. The jurors understood their struggle and how hard it must be for someone who had lived a life of crime to decide to finally turn his life around and, for once, help the good guys get the bad guys. They felt sympathy for the snitch, or at least wanted to rally behind him and his efforts at self-improvement. And thus they believed him. Or so we hoped. Since we almost always got our convictions, it seemed to work.
But snitching is a business. It’s cutting deals and negotiating to improve your lot. When I was a prosecutor, a defense attorney I frequently went up against was indicted for selling snitch information. For example, if he had a client who could help the feds on three cases, he knew that two cases would be enough to get the guy immunity, or at least get him some serious leniency come sentencing time. So he would, according to the charges, collect information from his client about the third case and sell it to another client who was empty-handed—who had no information to share. That way, both of his clients would have something to tell the feds about, and to testify about, to get leniency. He would allegedly split the profits with the client who sold him the information about the third case.
The annals of snitch lore are filled with similar stories. A chapter called “Snitch” in Jim Dwyer, Barry Scheck, and Peter Neufeld’s book, Actual Innocence, is filled with colorful anecdotes of wrongful convictions caused by incentivized witnesses who, once a new inmate was brought into the jail awaiting trial, would read newspapers to learn details about that inmate’s case so they could call the prosecutors and falsely claim that the new inmate had confessed to them. The confessions would appear reliable, because they would contain specific details about the crime that the snitch gleaned from the newspapers. One inmate was even called the “Snitch Professor,” because not only did he do it himself time and time again, but he taught other inmates how to play the game to their benefit.
My client Dewey Jones was wrongfully convicted based on snitch testimony. After he was arrested for murder and put in jail to await trial, another inmate named Willie Caton came forward and claimed Jones confessed to him while they were in jail together. Jones was exonerated, after spending nineteen years in prison, when DNA test results showed not only that he was innocent but that Caton was a liar. After Caton had testified in Jones’s trial, he became infamous in his region for his ability to repeatedly avoid prison sentences by cutting deals with police and prosecutors in exchange for testimony in criminal cases; he would dream up “confessions” by whomever the local law enforcement was hot to prosecute. This pattern of committing crimes and then testifying in exchange for leniency ultimately ended tragically in a car chase with police that culminated in a crash and shootout, the deaths of two young children, and the death of Caton himself. Caton was out on the streets in part because he had learned to work the system, and he wrongfully convicted at least one innocent person in the process.
Some cases can be solved only through the use of incentivized witnesses. That is understood. Thus, I don’t agree with the suggestions of some scholars that the use of incentivized witnesses should be abolished. Rather, it must be more carefully controlled and regulated. Ninth Circuit Court of Appeals Judge Alex Kozinski, among others, has called for stricter controls.55 In 1990, he noted, Los Angeles County instituted a new rule requiring a committee to screen such witnesses for objective signs of reliability before they could be used in court.56 This resulted in a steep decline in the number of cases where such witnesses were used, which suggests that many unreliable snitches were weeded out.57
Our system’s reliance on snitch testimony developed because of the intense pressure police and prosecutors feel to obtain convictions. And unreliable snitch testimony is often used in specific cases because of tunnel vision or even noble-cause corruption. Law enforcement believes that having a criminal testify against the suspect, and putting on a little show for the jury about why this criminal should be believed, helps get a bad guy off the street. It’s a good thing, furthering a good cause.
But to anyone who knows the full story of how we currently use incentivized witnesses, and understands how this practice really works behind closed doors, it should seem quite unsavory and prone to fabricated testimony. But those in the system don’t see it that way, not only because of tunnel vision but because it’s been done that way so long that it’s just accepted. It’s like a stain on the furniture that’s been there so long you no longer notice it. But there’s no objectivity or reliability in the process as it currently operates.
At present, forensic testimony is often unreliable because of confirmation bias and underlying methods that have not been properly validated. Prosecutors tell state-paid experts what result they need, and studies, as well as a plethora of wrongful conviction cases, repeatedly have shown that this confirmation bias leads to inaccurate results tilted toward the prosecution. But the problem is deeper than simple confirmation bias. Entire subdisciplines of forensics are systemically corrupt, even in cases where no such bias exists.
As with eyewitness identification evidence, confession evidence, and snitch testimony, the current problems with forensics are so deep and complicated that it would take an entire book to do them justice. And books have been devoted exclusively to the current problems with forensics and to these other forms of evidence.58 There’s a wealth of literature on forensics. In short, most types of forensic disciplines that we have used in the past and continue to use today in the criminal justice system are not based on scientific principles with a known error rate. For years we have been convicting people in this country based upon “expert” bite mark testimony (an expert claiming the suspect’s teeth match the bite mark on the victim’s body), shoe-pattern matching testimony (an expert claiming the shoe print left in the mud at the crime scene matches the defendant’s shoe print), gunshot residue testimony (experts claiming gunshot reside was found on the suspect’s clothing after a shooting), tire tread matching testimony, bullet fingerprinting testimony (experts claiming the bullet from the crime scene came from the suspect’s gun), friction ridge analysis in fingerprint testimony, and many other so-called forensic “sciences.”
With the exception of DNA testing, none of the disciplines used in forensics today was developed in an objective setting through a proper validation process with a known error rate. Experts often testify to such things as that the crowbar found in the defendant’s house—and only the defendant’s crowbar, to the exclusion of all other crowbars in the world—could have made the particular marks on the victim’s doorframe. I have an Ohio Innocence Project case right now where that exact claim was made to convict my client. And that seems like very incriminating evidence indeed. But how can one testify with absolute certainty that the defendant’s crowbar made the scratches on the doorframe in question when all other crowbars in the world look pretty much the same? No one can answer that question. There were never studies performed to show an accuracy or error rate, or to prove that such claims are even defensible. And we now know that such claims can’t be justified, because many people who were later exonerated by DNA testing were originally convicted on such exaggerated, unscientific “expert” testimony. Also, the research of psychologists like Itiel Dror shows us that these “experts” will change their answers at alarming rates—for example, from a “match” to “no match”—depending on what they are led to believe the “right” answer is before they start.
It’s not just me saying this. In 2009, as discussed earlier, the National Academy of Sciences came out with an alarming report on forensic science in the United States, one that called into question the basis for nearly every forensic discipline—the only one that passed muster was DNA testing—and recommended that Congress create a National Institute of Forensic Science to help mitigate the problem. The NAS wanted the NIFS to create guidelines and rules to ensure that all forensic testimony in the courtroom is actually based on science with a known error rate, and that standards and accreditation are employed to ensure the accuracy and reliability of all such evidence. It recommended that all forensic crime labs be removed from the control of law enforcement agencies and prosecutors’ offices and be made truly independent, and that methods be developed to eliminate the confirmation bias that currently pervades these fields. It proposed corrective measures that might include, for example, providing fingerprint samples from six different people, instead of just the suspect’s fingerprint, to an expert who is trying to determine whether there is a match to the fingerprint from the crime scene, and removing the expert’s bias by not revealing the “right” answer.
But, to date, the NAS’s thirteen recommendations have been by and large ignored.59 Most forensic disciplines are still systemically corrupt, and many are still being used in the courtroom. They are too often the products of tunnel vision and bias, masquerading to jurors as infallible, awe-inspiring wonders as depicted on television shows like CSI, on par with putting a man on the moon.
To use another example from the documentary Making a Murderer: part of Steven Avery’s defense to the charge that he murdered Teresa Halbach was that Avery’s blood found in Halbach’s car had been planted by the police. It’s conceivable that the police could have planted it, because they had Avery’s blood samples in vials from his prior wrongful conviction case. To counter this defense, the prosecution sent the blood found in Halbach’s car to the FBI to test for EDTA, which is a preservative that would have been present in Avery’s blood contained in the vials. If no EDTA was detected in the blood found in the car, the prosecution surmised, that would prove the blood didn’t come from the vials. Defense attorney Jerry Buting quipped when he learned of this plan, “I don’t trust the FBI at all. And I think that they’re going to come up with some dishonest test that somehow claims that the blood in the vial is different than what was found at the scene.”60
I laughed out loud when I saw that scene, because I know how Buting feels. Like me, he’s been engaged in this type of work so long that he knows how forensics are often used by prosecutors as a means to an end. And Buting’s fears may have been founded. The FBI originally said it would take months to run such a test, because these tests had been abandoned years earlier due to concerns about their reliability, and thus a new testing mechanism would have to be created specifically for that case. But when the prosecution needed results immediately, the FBI was somehow miraculously able in a matter of a few weeks to run the test and get a result that favored the prosecution. And at trial, Buting pointed out that the FBI lab analyst was told by the prosecution what answer he was supposed to arrive at before he created the test. The evidence submission sheet stated the purpose of the test was to “eliminate the allegation” that the police had planted blood.
A defense expert at Avery’s trial testified that the FBI’s test was unreliable. (Indeed, the results of such a test had never before been admitted in court, when contested, out of unreliability concerns.) Furthermore, the defense expert noted that the FBI’s results could not be subjected to scrutiny, because the FBI had refused to turn over the details and processes of the new test it had created just for that particular case. But juries usually don’t get these nuances. They see an impressive FBI analyst who flew in from Washington DC to testify, and typically accept such testimony as the gospel. I have previously described this phenomenon as the “Reverse-CSI Effect.”61 And that’s what may have happened in Avery’s case, just as Buting feared.
And the FBI blood expert in Avery’s case even gave an opinion on several samples of blood he hadn’t tested! He stated that in his expert opinion those samples did not contain EDTA, meaning they could not have been planted. This statement caught Buting by surprise, causing him to ask the expert to repeat it. When the FBI expert repeated his statement, Buting commented, “OK, I wanted to know how far you were willing to go.”
Tunnel vision can produce a snowball effect in cases of wrongful conviction, where an initial, single piece of errant evidence gives rise to what appears to be an overwhelming case by the time of trial.62 If an eyewitness mistakenly identifies an innocent person from a photo lineup, for example, the police may become even more convinced that they have the right suspect, leading them to put pressure on the forensic experts to find a fingerprint match with a smeared latent print left at the crime scene. After the expert then gives them the fingerprint match they were looking for, they may then—even more certain of the suspect’s guilt than before—interview potential witnesses in the case and “suggest” additions or corrections to their statements to make the overall picture more compatible with the suspect’s guilt. And with enough pointed questions, a witness or two might start “remembering” things the way the police want them to.
Now that the police have both an eyewitness and a fingerprint match, and other witnesses providing incriminating details that undermine the suspect’s alibi, they may approach inmates in custody with that suspect and ask if any of them by chance overheard the suspect talk about committing the crime. With inmates eager to work time off their own sentences, it’s often not hard for the police to find someone who is more than willing to concoct a story about overhearing the suspect talk about committing the crime. Or they may go right to the horse’s mouth and elicit a false confession from the suspect himself, pressuring him with the facts that an eyewitness has already identified him and that his fingerprint has been found at the scene, making it appear that a quick confession and a plea for leniency is his only way to avoid the death penalty.
By the end, the case against the innocent suspect appears overwhelming. But each piece of evidence was created from the ones before it in a “feedback loop,”63 and each can be traced back to the initial piece of errant evidence—the initial snowflake—that has become a giant snowball by the end.