Khemwatie Bedessie, a 39-year-old immigrant woman in New York City, was convicted last year of raping a 4-year-old at a daycare center in Queens, though the facts of the case strongly suggest she is innocent. Her conviction resulted solely from a confession, which she says is false and was coerced from her by a detective.
In the 1930s, the Supreme Court outlawed “the third degree” during police questioning. Interrogators can no longer beat people, keep them awake for days, or threaten them with death to get a confession. Rogue behavior still surfaces. Chicago is still investigating a police district that routinely applied electric shocks to suspects less than a generation ago. But this isn’t the Depression Era, and coercive interrogations are no longer supposed to be allowed.
It’s not the 1980s, either. That decade marked the eruption of the McMartin Preschool case, in which several California childcare workers, among them elderly women, were accused of most bizarre and extreme sex abuse against children. McMartin, with its claims of mutilated rabbits and sodomy in underground tunnels, turned into the longest and most expensive criminal case in U.S. history, before it collapsed in 1990, with acquittals and hung juries. Dozens of copycat cases from the same period have since been debunked, and today child protection authorities tell us they know child sex abuse investigations can go haywire, but they have ways to keep them on track so people aren’t treated unjustly.
Even so, Khemwatie Bedessie was accused and convicted without any substantial evidence, except for her confession. Was it really coerced and false, as she claims? We’ll probably never know for sure because police didn’t record the interrogation that led to her self-incriminating statements. Lack of recording is one reason Bedessie deserves the benefit of the doubt. Her interrogation should have been videotaped, just as all questioning should be when people are detained during investigation of serious crimes. Among law enforcement agencies around the country, videotaping is catching on, and that’s laudable. But even if taping becomes universal, it won’t come near to eliminating false convictions based on false confessions. To make a real dent in the problem, we need to first recognize that when it comes to investigating crimes, we’re still in the epoch of the Inquisition.
Bedessie case is instructive, and it has a back story. She is one of nine siblings from Guyana, and grew up very poor there. At age 3, she was kicked in the head by a donkey; after that she suffered bouts of writhing and foaming at the mouth, which her family calls “seizures” or “anxiety attacks.” She never received medical treatment for them, and because classmates teased her about the attacks she dropped out of school after fifth grade. She cannot add or subtract small numbers, and her writing looks like a 7-year-old’s. After coming to the United States five years ago, she lived with her mother and worked 11-hour shifts, doing cleaning at a small daycare center in Queens. There she was known by the children as “Teacher” and by their parents as “Anita.”
One preschooler was a boy I will call Sam. At Bedessie’s trial this spring, Sam’s mother testified that when she first put him in daycare at age two so she could take a job, she was anxious about leaving him. Soon she started asking him if anyone there was sexually abusing him. She asked randomly and frequently. “No, mommy,” Sam always replied.
Then, one day in winter 2006, Sam developed a fever and a rash on his buttocks. At the doctor’s, he was diagnosed with flu. But his mother, again, felt worried. Again, she asked him about abuse. This time Sam, now 4, said “yes.” Taken to a hospital, he told a nurse he’d been raped by “Anita” – not his name for Bedessie but his mother’s. A police officer was called, but Sam would not repeat the statement. And medical personnel did not change their diagnosis of the rash. They still made no finding that it was caused by sexual abuse.
That left nothing except a preschooler’s word – which was spotty, and could have been tainted by his mother’s constant questions. And there was another problem with the case: it is astronomically rare for females as old as Bedessie to commit sex crimes against tiny children. Given this fact, what is the probability that the rape of a 4-year-old by a middle-aged woman would be discovered purely by accident, by questioning a child whose original complaint – which triggered the questions to begin with – had nothing to do with sex abuse? The likelihood is miniscule. The most probable explanation for Sam’s allegation of rape is that it was false, evoked by his mother’s fears and the boy’s suggestibility.
Not surprisingly, the detective in charge of the case, Ivan Borbon, was getting nowhere after a week of investigating. But instead of calling it quits, he decided to bring Bedessie in for questioning. Wearing plain clothes and driving an unmarked car, Borbon arrived at the day care at 9 a.m. one day. Bedessie said she thought he was a child protection worker. Borbon did not alert her to the misconception, and he told her they were going to his “child protection” office. It turned out to be a police interrogation room. There, Bedessie later testified, Borbon began cursing at her and calling her a child molester. He displayed a tape recorder and said he’d “wired” Sam. He claimed he had, on tape, the sounds of Bedessie forcing the child to have intercourse with her in the daycare bathroom. Incredulous, she asked him to play the tape. He refused, cursed some more, and said Bedessie had two choices. She could say then and there that she had raped Sam and she would be released to go home. Or – as she put it at trial – she could continue to profess innocence and “go to Rikers and never see my mommy” again.
“I do whatever he tell me to do,” Bedessie later testified. She says she has no memory of confessing (family members say she dissociates when she has her “anxiety attacks”).
But she did make a confession, after only three hours in custody. It was videotaped. In her statement, she responds to questioning by describing being fully penetrated sexually, for several minutes, on a toilet, by preschooler Sam. She characterizes the penis of this 4-year-old as being as long as a ballpoint pen, and of “about two inch thickness.” She speaks a notably creolized English, and it is not clear she understands everything she is asked. At trial a year later, she said she did not know the meaning of the words “masturbation,” “stroking,” “orgasm” or “immoral.”
Bedessie’s attorneys tried to put a witness on the stand: Richard Ofshe, an internationally recognized expert in false confessions. The judge would not allow it. He said the jury could make up its own mind about the veracity of Bedessie’s incriminating videotape. After only a couple of hours’ deliberation, they convicted her.
Though Ofshe did not testify, he watched Bedessie’s confession and interviewed her before her trial. He finds her account of coercion very credible, and says many people make false confessions after much less time than the three hours it took for Bedessie to begin her statement. Her description of the interrogation, Ofshe says, sounds like many others he has heard, in which evidence later surfaced to show that the defendant was innocent, even though he or she had earlier confessed. Ofshe and every other researcher who has studied false confessions note that they are easily extracted by interrogators. That’s because of how interrogation works – even when it’s done legally.
The Arizona v. Miranda decision, with its caveats about the right to stay silent and its offers of lawyers, was issued by the Supreme Court in 1966. Since then, legal police questioning supposedly has dispensed with 24/7 marathons and physical assault. Now, interrogations concentrate on psychology. But even when everything is on the up and up , questioning in detention is no tea party. According to the law, cops can get people to talk by yelling, insulting them, invading their personal space, saying there’s evidence when there isn’t, and feigning sympathy about the crime (“After all, she was dressed like a slut. I know she was asking for it, huh?”).
A widely used training manual recommends that the interrogator physically crowd up next to the suspect and insist he or she is guilty, cutting off any bodily or verbal protestation of innocence. “The interrogator must rely on an oppressive atmosphere of dogged persistence,” advises the manual, “leaving the subject no prospect of surcease. He must dominate the subject and overwhelm him.” These techniques “suggest that only confession will bring interrogation to an end.” In this way, the manual instructs, it is possible “to induce the suspect to talk without resorting to duress or coercion.”
But, at some point on the continuum of trickery, duress and threats, cops can step over a line. The resulting confession is what most people think of when they read reports from organizations such as the Innocence Project. According to that group, in over of quarter of DNA exonerations, innocent defendants pleaded guilty or made false confessions. Many such confessions and pleas were produced because police officers promised leniency at sentencing in exchange for a confession. Such deals are not allowed. Or the interrogator threatened bodily harm, warning the suspect, for instance, that confessing would be the only way to avoid the death penalty. (Bedessie says that Borbon, the detective who interrogated her, told her about the terrible treatment accused child molesters get at Rikers. He said she could avoid going there by confessing).
According to a raft of social science and psychology research done over the past two decades, techniques like these are especially likely to produce false confessions when used on juveniles, the mentally ill, the poorly schooled, immigrants, and those with impaired cognition (Bedessie fits at least two of these categories).
It’s also agreed that illegal practices occur frequently in the interrogation room, and that cops later lie about them on the stand. And when there is an argument about veracity, research suggests that no group of people – not judges, prosecutors or juries – can tell whether a confession is true or false simply by reading a transcript or watching the video. That is why not just the confession should be recorded, but also the full interrogation that led up to it. The idea is to avoid methods that – as the Supreme Court has put it – “shock the conscience” and “offend the community’s sense of fair play and decency.”
Ten years ago, only two states were recording interrogations. Now, nine states and the District of Columbia do, and they are joined by more than 500 local police departments nationwide (some record only for murder cases, others for lesser felonies as well). Increasingly, taping is the trend. It’s spreading relatively slowly, but it’s spreading, says Northwestern University legal scholar Steven Drizin, an expert on false confessions who has advocated for taping for years. He thinks the scales would really tip if federal agencies started making recordings.
So far, the feds have said “no.” But last year, media eyebrows were raised when the DOJ released documents related to how eight U.S. Attorneys were fired under former Attorney General Alberto Gonzales’ watch. Speculation is that one of the fired attorneys, Paul Charlton, in Arizona, was let go because he was investigating Republican Congressman Rick Renzi, a Bush loyalist, about a 2005 real estate deal. Either that or Charlton angered the DOJ for not prosecuting enough obscenity cases based on adult porn. Gonzales’ office demurred, saying that a major reason Charlton was canned was that he wanted to start a pilot project for the FBI and other federal agencies to start experimenting with videotaped interrogations. When the documents came out, one of them – from the FBI – objected to Charlton’s idea and commented that “as all experienced investigators and prosecutors know, perfectly lawful and acceptable interviewing techniques do not always come across in recorded fashion to lay persons as a proper means of obtaining information from defendants.” More pointedly, the memo mentioned worries that jurors could find “proper interrogation techniques unsettling.”
Couple these anxieties with steady media attention to the problem of false confessions, and it might seem odd that judges, juries, and the public in general still find it so hard to believe that someone like Khemwatie Bedessie would say she was guilty if she wasn’t. Inside and outside the courtroom, what is the problem?
The most proximate answer is that, logistically speaking, the U.S. is heavily invested in a criminal justice system that would be paralyzed without confessions. Ninety-two per cent of felony convictions are obtained by plea bargains or confessions. That’s a far higher rate than in other countries (Italy’s, for example, is 8 per cent, and Norway doesn’t allow plea bargaining at all).
Relying on confessions to prosecute crimes is thrifty because it avoids the need for costly investigations. But it’s also very destructive to justice, according to Jerusalem University criminologist Boaz Sangero. Writing in a recent issue of Cardozo Law Review, he lists several problems. The first is that, after a suspect is apprehended, police tend to ignore serious investigation; instead, they focus on getting a confession. And once the confession is obtained, any other work going on at all typically ends. The push to handle cases this way encourages misbehavior in the interrogation room.
Further, reliance on confessions promotes disgraceful conditions of detention. Jails are often worse than prisons. Filth, bad food, lack of sunlight, crowding and violence pressure people to say they did something – anything, whether it’s true or not – just to get out of lockup. Then, because they’ve confessed, we figure it’s OK to keep others like them in awful cells – and to bring in more detainees for interrogation. It’s a vicious circle, and most who get trapped in it are poor, uneducated, and unacculturated. Their marginal status is bound up with the moralistic judgment that they are different from us, and therefore bad. Their badness reinforces our willingness to keep a bad system in place. It probably also allows us to export illegal interrogation – our 1930s-era torture, updated – to places like Abu Ghraib and Guantanamo.
Beyond fear of the bad “other” and desire for a bargain, though, there’s a more fundamental, existential reason why dependence on self-incrimination is mean and unfair. As Sangero notes, any kind of interrogation which focuses on obtaining confessions – legal or illegal – probably violates people’s rights. That’s because, from the point of view of self-interest, confession makes no sense at all. People are asked to help themselves by condemning themselves. It is deeply irrational.
That irrationality is especially apparent in the many confessions made, even though they were not extracted directly by police questioning. In fact, as Sanjero notes, it’s possible that most confessions arise not from external coercion but from states of dependency and abjection that people internalized before they were ever interrogated.
Historical and legal records abound with examples. After Charles Lindbergh’s baby was abducted, over 200 people walked into police stations and said they were the kidnapper. More than 30 told authorities they were the murderer of a woman who came to be known as “The Black Dahlia” – a Hollywood actress whose mutilated body was found in a vacant lot in Los Angeles in the 1940s. In a case that truly smacks of internalized abjection and desire for quick death, Heinrich Himmler lost his pipe while visiting a concentration camp during World War II. A search ensued, but on returning to his car Himmler found the pipe on his seat. Meanwhile, the camp commandant reported that six prisoners had already confessed to stealing it.
Since they are not products of police interrogation, no amount of videotaping will eradicate these confessions. Yet, we accept them. At least partly, this is because quick admissions of guilt are cheap, and easy on the justice system. But, more fundamentally, the very concept of confession is deeply embedded in our culture.
It was not always so. Ancient Jewish law barred criminal confessions. In Talmudic commentary – cited in the Supreme Court’s Miranda decision, by the way – the rabbinical scholar Maimonides notes, “The court shall not put a man to death or flog him on his own admission.” Additional evidence and witnesses are needed, Maimonides explains, because the impulse to confess is, by definition, self-destructive. Of a man who professes guilt, there is always the possibility that he is “one of those who are in misery, bitter in soul, who long for death …perhaps this was the reason that prompted him to confess to a crime he had not committed, in order that he be put to death.”
Since the 1551 Council of Trent, however, the Roman Catholic Church has taught that confession is good for the soul – yea, even necessary, to save it and purge it of impurity. This religious notion has since been incorporated into law and into the modern, secular definition of the self. Being a fully realized person today requires full disclosure to family, friends, and even (in the case of writers, artists and public figures) to the polity: of one’s deepest emotions, darkest sexual impulses, and past misdoings. Confession isn’t just good for the self. We need confession to be a self.
But when self meets soul in the modern justice system, it’s a train wreck of contradiction. As Yale University comparative literature scholar Peter Brooks notes in his book Troubling Confessions, “That we continue to encourage the police to obtain confessions whenever possible implies a nearly Dostoevskian model of the criminal suspect … we want him to break down and confess, we want and need his abjection since this is the best guarantee that he needs punishment, and that in punishing him our consciences are clear.” On the other hand, our Mirandan insistence “that the suspect’s will must not be overborne, that he be a conscious agent of his undoing, of course implies the opposite, that we don’t want Dostoevskian groveling in the interrogation room, but the voluntary (manly?) assumption of guilt. Hence the paradox of the confession that must be called voluntary while everything conduces to assure that it is not.”
It wasn’t so long ago that masters of American jurisprudence were actively grappling with this contradiction. In the 1966 Miranda decision, Earl Warren recommended that the police find other evidence to solve a crime than the “cruel, simple expedient of compelling it from [the suspect’s] own mouth.” Twelve years before Warren made that statement, Abe Fortas, who later would replace Warren on the Supreme Court, wrote that “Mea culpa belongs to a man and his God. It is a plea that cannot be exacted from free men by human authority.”
Today, Sangero agrees with these liberal lawmakers from a bygone era. He wholly opposes the eliciting and use of confession to solve and prosecute crimes. But, if confession isemployed, he believes the case should never go forward unless meaningful evidence is first gathered from sources independent of the confession – evidence that strongly shows, rather than merely suggests, that the suspect committed the crime. Many people fear that such a policy would allow lots of guilty people to go free. Sangero dismisses their worries. Forensic science in the U.S. today is so sophisticated and high tech, he says, that police have only to use it. All that is required to convict criminals justly is that the cops do their job.
Sangero is very leery of putting too much emphasis on recording. Sure, he says, it’s needed. But narrowly focusing on videotaping reforms does not encourage the police to redirect investigations away from defendants’ self-incrimination and toward the gathering of independent evidence. Obsession with recording can encourage practices such as “non-detentive interviewing.” It’s an increasingly common ploy, in which suspects are seduced into chatting – as Bedessie was when she was visited by the supposed “child protection worker,” who turned out to be a policeman – without being read their Miranda rights. Only after the car door is locked, the drive has begun, and the interrogation room is sighted, does the suspect get officially detained and put before a camera. By then, for someone like Bedessie, it may well be too late to take exercise one’s Miranda rights.
Bedessie is now in the first year of a 25-year prison sentence. Her post-conviction legal work is being done by prominent Manhattan attorney Ron Kuby. He believes she has a good shot at having her conviction overturned because of the trial judge not letting the jury hear expert testimony about false convictions. Nowadays, that’s solid grounds for appeal, and even the assistant DA who prosecuted the case knows it. Pretrial, she advised the judge that it wouldn’t hurt the state’s case to let the defense put on a witness to warn jurors that Bedessie might have falsely incriminated herself. It wouldn’t matter because the confession spoke for itself. And no jury would think otherwise.
EMILY HOROWITZ is a professor of sociology and criminal justice at St. Francis College (Brooklyn, NY). She serves as a director of the National Center for Reason and Justice (www.ncrj.org), an innocence project for people wrongly accused or convicted of crimes against children and a sponsor of Khemwatie Bedessie. She can be reached at firstname.lastname@example.org