The Supreme Court Can Substantially Reduce the Number of False Confessions
False confessions are a criminal justice system failure that the Supreme Court should have—and could have—fixed long ago
This piece is the third in a series on defects in our criminal justice system that lead to wrongful convictions. The first piece introduces these seven critical breakdowns in the rules of constitutional criminal procedure, and the second examines the problem of eyewitness misidentification.
In the middle of the night on November 12, 1996, a K-Mart cashier in Sterling Heights, Michigan walked to her car in the store parking lot after finishing her shift. As she entered the vehicle, a man claiming to have a gun forced her to slide into the passenger seat. He demanded the keys and drove, and he threatened to kill her if she looked at his face. He found an isolated spot, pulled the car over and raped her. Finally, he released her, and she ran to a local business and called the police.
Three days later, officers found a 17-year-old named Nathaniel Hatchett, who looked vaguely like the victim’s description of the attacker—same race, but younger and taller—driving the victim’s stolen car. (How Hatchett found his way into the victim’s car the day he was arrested remains unclear.) Hatchett was arrested, brought to a police station and interrogated. Though he claimed to have had nothing to do with the kidnapping or rape, police grilled the terrified and confused teen, whose IQ was only 74, for hours. They promised that if he complied, he could go home. Unaware that he was being manipulated and unable to understand that he was sealing his own doom, Hatchett took responsibility for the crimes. Two DNA tests on the rape kit evidence showed conclusively that Hatchett was innocent. Yet he spent 11 years in prison, before he was finally exonerated and freed in 2008 (I worked on his case as a law student intern at the Cooley Innocence Project at Western Michigan University).
Regarding forced false confessions like Hatchett’s, the DNA revolution in criminal forensics has shown that suppressing exculpatory evidence is a common law enforcement practice. Hatchett’s “confession” consisted of information fed to him by investigators. These same investigators later emphasized Hatchett’s guilt from the witness stand by claiming he’d disclosed facts about the crime only the perpetrator could have known. All of the learned judges who reviewed Hatchett's case expressed perfect confidence in the reliability of his confession. Had the Cooley Innocence Project, who engineered the second DNA test publicly affirming the exonerating results of the first, not later taken Hatchett’s case, he would likely still be buried in the bowels of a penitentiary serving his 25-to-40-year sentence.
The National Registry of Exonerations reports that there have been 388 exonerees who falsely confessed to crimes since 1989. But based on the extraordinary efforts, luck and sometimes both required to achieve these exonerations, it’s reasonable to assume that this is just the tip of the iceberg. And false confessions that lead to wrongful convictions are a problem as old as policing itself. What makes them especially frustrating, however, is that, unlike the other major causes of wrongful convictions, the Supreme Court announced a formula that would have dramatically reduced this problem decades ago—only to quickly reverse course and ensure its continuation.
A Coercion-Free Interrogation Room?
With his famous opinion in Miranda v. Arizona in 1966, Chief Justice Earl Warren promulgated a set of rules designed to create a coercion-free interrogation room. However, almost immediately after Warren’s retirement in 1969, the Supreme Court launched an effort, still underway, to restore and even enhance police powers during questioning. While never completely overturned—and even occasionally affirmed—over the past 56 years, Miranda has died a death of a thousand cuts, the most recent coming earlier this year when the court ruled that suspects are barred from suing the police for violating their Miranda rights. It’s as if the court administered us medicine only to instantly begin pumping our stomachs—and the result has been that cases like Nathaniel Hatchett’s continue to recur.
While it may be hard for some of us to imagine falsely incriminating ourselves to police, the Innocence Project reports that since its founding in 1992, 29% of DNA exonerees falsely confessed to the crimes—mostly rapes and murders—for which they were convicted.
Nearly everyone has a physical and psychological breaking point at which their ability to resist an interrogator is overborne. Confronted with enough pain, shame or fear, most of us will admit to whatever’s necessary to get just a temporary respite, even knowing the result may be years in prison or a death sentence. And skilled police interrogators are adept at weakening our will and finding our breaking points.
Police often find the temptation to cross the line from lawful persuasion to unconstitutional coercion irresistible because confessions are especially apt to doom defendants at trial. In most cases, even otherwise sympathetic jurors blanch at acquitting a defendant after reading or hearing the kind of confession detectives are trained to extract—confessions designed to maximize culpability by including the specific motives and grisly, granular details of the crimes perpetrated. As a lawyer arguing on behalf of the National District Attorneys Association explained during oral arguments in the Miranda case, an admission of guilt is “the most important piece of evidence in every case that [police] bring before a court of justice.” An incriminating confession is a windfall benefit that makes a case easier for the state in every way, often sparing it the bother of future investigation and, by strengthening plea negotiations, sometimes obviating trials altogether.
The Court’s Problem To Solve
Miranda wasn’t our criminal justice system’s first time weighing in on false confessions. From the mid-18th century, English judges recognized the unreliability of coerced confessions and banned them from criminal trials. The Supreme Court heard fewer criminal cases in its early history, but endorsed this approach when it first considered the issue in 1884 in Hopt v. Utah. The court had wrongful convictions in mind when it considered the case, recognizing that although voluntary confessions carry great weight because “one who is innocent will not imperil his safety or prejudice his interests by an untrue statement,” a confession’s trustworthiness “ceases when the confession appears to have been made either ... because of a threat or promise by or in the presence of such person, which, operating upon the fears or hopes of the accused, in reference to the charge, deprive him of that freedom of will or self-control ...” The court couldn’t have been more clear: Involuntary confessions can’t be trusted and interfere with the truth-seeking process of trials. Because they may steer juries toward an erroneous guilty verdict, they mustn’t be admitted into evidence.
However, the court’s opinion in Hopt did little to curb the routine use of physical and mental brutality—known as the “Third Degree”—in American police stations. Due to principles of federalism and states’ rights, early Supreme Court rulings on this issue applied only to federal cases whereas the vast majority of criminal cases are investigated and tried at the state level. It wasn’t until 1936 with Brown v. Mississippi, which involved an interrogation that recalled the horrors of the medieval period, that the Supreme Court began using the U.S. Constitution—more particularly, the Fourteenth Amendment’s due process clause—to regulate state and local police interrogations.
In Brown, a sheriff’s deputy, with the help of a group of local vigilantes, whipped murder suspects and subjected one to a sadistic mock execution by stringing him up by the neck—during a time when lynching was all too common in the South—and only agreeing to end this torture if he confessed to the crime. The court, sickened by these facts, ruled that convictions based on plainly coerced, and therefore unreliable, confessions like these violated the Fourteenth Amendment’s promise of a fundamentally fair trial in state courts. It reasoned that by using coerced confessions, the state had “contrived” a conviction and reduced the trial to a mere “pretense.”
The court’s ruling came on the heels of the publication of the “Wickersham Report,” commissioned by President Herbert Hoover, which announced the embarrassing truth that “the inflicting of pain, physical or mental, to extract confessions ... is widespread throughout the country.” Brown was an unprecedented intervention into state and local criminal procedure. The Third Degree was now the Supreme Court’s problem to solve.
Taking Police Misconduct Seriously
The thirty years between Brown and Miranda were frustrating ones for a court that, for the most part, took its responsibility to combat this police misconduct seriously. But the problem quickly proved daunting and unwieldy: With Brown, the court began applying a “totality of the circumstances test” to distinguish voluntary from involuntary confessions, according to which it examined all of the relevant facts of an interrogation to determine whether police had extracted the confession by overriding a suspect’s freewill. Under this test, the justices would measure the pressure applied by police against the suspect’s subjective ability to resist to determine whether the latter’s due process rights had been violated. While perhaps workable in theory, the test proved subjective and arbitrary—often hopelessly so—in practice.
Part of the problem was that interrogations were frequently unrecorded and conducted in secret. Only police and the accused were witnesses, and either side might have a strong incentive to later lie in court. The justices were often left maddeningly uncertain of—yet obliged to rule on—the “totality of the circumstances” that existed in the interrogation room. To ameliorate this problem, many states have adopted mandatory recording laws.
After struggling with these cases for 13 years as chief justice, Warren had had enough. He’d spent his entire career in law enforcement—as a district attorney, an attorney general and a governor—so he knew through personal experience what went on in interrogation rooms. With the exception of his participation in the persecution of citizens and immigrants of Japanese ancestry during World War II, which he came to profoundly regret, Warren spent decades striving to build a law enforcement culture guided by moral decency and subject to the rule of law. In his opinion in Spano v. New York, a death penalty case based on a forced confession, he wrote: “[T]he police must obey the law while enforcing the law ... liberty can be as much endangered from illegal methods used to convict those thought to be criminals as from the actual criminals themselves.”
The brutality and bullying that police showed toward their helpless captives—many of them uneducated, poor or otherwise vulnerable—and their self-exonerating deceptions afterwards, repulsed Warren’s strong moral core. While accused of serious crimes, these suspects remained citizens entitled to a presumption of innocence and the benefit of law. The boldness of the civil libertarianism in his Miranda opinion, combined with the profound distrust of police it contains, was a kettle that had boiled over.
Miranda announces, as a matter of constitutional law, that the extreme power disparity between police officers and suspects during custodial interrogations creates an inherently coercive atmosphere—one so intimidating that courts should presume that, without some leveling of the power imbalance, suspects are incapable of asserting their Fifth Amendment privilege against self-incrimination. To mitigate the coercive pressure on suspects, police are required to recite the famous Miranda warnings we all know from TV crime dramas and police procedurals. For incriminating statements to be admissible in court, the rights contained in those warnings, including the right to remain silent and have a lawyer present during questioning, must be understood and voluntarily waived prior to questioning. And courts would not simply be taking law enforcement at its word that a suspect waived his Miranda rights: Instead, the state has the burden of proof to show that the suspect freely and knowingly chose to give up his rights. Under this approach, when police failed to inform the accused of his Miranda rights, there became no need for the court to apply the totality of the circumstances test: The confession would automatically be inadmissible.
Typical of Warren, his Miranda opinion is written as a moral indictment against unlawful police practices. He refers readers to a series of post-Brown cases involving burning, beating and other forms of physical and mental abuse that highlight the ineffectiveness of the voluntariness test. In footnote 24 of his opinion, he cites the famous “Career Girls” false confession case, in which an innocent man was made to take responsibility for the brutal rape of one woman and murder of two in New York.
He further explains that even the prevailing lawful methods police use to extract confessions—which include playing good cop/bad cop, lying about the strength of the evidence, giving false legal advice and other forms of trickery and pressure—are indifferent to truthful fact-finding. By the time suspects are in custody and interrogated, police have usually developed their theory of the crime and are convinced of the suspect’s guilt. The purpose of the interrogation isn’t to get answers: The purpose is to get suspects to agree to their answers. Cops are no longer investigating: They’re manipulating. Protestations of innocence are ignored and redirected. For the interrogator, anything short of confirmation and self-condemnation is failure.
The allure of these techniques is that they’re effective. The problem is that they’re so effective that sometimes they work on the wholly innocent. For these suspects, the experience can be a Kafkaesque nightmare. Miranda was designed to revolutionize this disparate power dynamic by allowing suspects to escape these conditions simply by saying “no.”
Dismantling Miranda—and the Rights of the Accused
The social upheaval and rise in crime during the late 1960s created an electoral opportunity for Warren’s old political rival from California, Richard Nixon, in 1968. Nixon expressly ran against the Warren court’s criminal procedure rulings in his run for presidency, describing them as “soft on crime” and promising to restore “law and order.”
None of these rulings was more controversial, or resulted in more conservative backlash, than Miranda. Nixon would appoint four justices in rapid succession, and in short order, the court began scaling back Miranda’s influence. The effort has grown since tough-on-crime Reagan, Bush and Trump appointees have occupied leading factions on the court.
Much of the criticism these justices have levelled against Miranda has been rooted in concerns for public safety that aren’t altogether irrational. They fear that Miranda will keep dangerous criminals on the street by either preventing confessions or keeping truthful confessions obtained without warnings from juries. While fair and understandable, these concerns are simply outweighed by the long, stubborn history of false confessions that lead to wrongful convictions. Moreover, some truthful confessions prevented when suspects exercise their rights can be made up for by additional detective work using legal evidence-gathering techniques. Those excluded from evidence because interrogators flout or forget Miranda’s requirements can be remedied by better police training. Both of the primary arguments that Miranda makes us less safe can be answered, to a large extent, by better policing.
In 1971, the court began dismantling Miranda by ruling that, while inadmissible in the prosecutor’s case-in-chief, prosecutors can impeach a defendant’s credibility when he takes the stand by confronting him with un-Mirandized prior inconsistent statements made to police. In 1974, the court ruled that the right to hear Miranda warnings isn’t constitutional per se, but a mere “prophylactic” rule—less binding and of a lower legal order—created to protect a defendant’s actual constitutional right not to be forced to speak under the Fifth Amendment. Having decided that Miranda rights were unrelated to the Constitution, the justices began limiting their force more freely. For instance, in a 1981 case involving a loose gun at a grocery store, the court ruled that the state’s interest in public safety permitted police to question an unwarned arrestee about the location of the gun. This “public safety” exception was cited by the Justice Department when federal agents interrogated Dzhokhar Tsarnaev, one of the Boston Marathon bombers, unwarned for 13 hours—denying his repeated requests for a lawyer. (Because the legally admissible evidence against Tsarnaev was overwhelming, the government never used his un-Mirandized statements against him at trial, so no judge has ruled on the government’s conduct. He was sentenced to death.) In 2004, the court ruled that physical evidence discovered as a result of an un-Mirandized statement could be used against the suspect at trial. In this case, the unwarned suspect admitted to possessing a gun. The gun was then found, and prosecutors used it against the suspect at trial.
But perhaps most devastating to the rights of the accused is the notorious “clear statement rule.” The court had long interpreted Miranda to mean that a suspect can cut off questioning at any time by “invoking” his rights to silence or counsel. In 1994, however, the court ruled that when a suspect interrupted his interrogation with “Maybe I should talk to a lawyer?,” he did not clearly enough express his wish to have counsel present. “Rather,” according to the court, “the suspect must unambiguously request counsel.”
The court extended the clear statement rule in 2010, this time applying it to the right to silence rather than counsel. The ruling in Berghuis v. Thompkins goes beyond the counterintuitive and into the surreal. After the detectives in this case informed Thompkins of his right to remain silent, Thompkins took detectives at their word and remained almost completely silent for over two hours. Finally, after police exploited his sense of religious shame by asking whether he prayed to God to forgive him for the murder, Thompkins broke and answered “yes.” The court ruled that actually remaining silent—upon police instruction that you have the right to do so—does not clearly express a desire to exercise your Miranda right to remain silent. Instead, the right to enjoy the Miranda right to silence is preconditioned on speaking up—and the police have no duty to warn suspects of this precondition.
The clear statement rule can also punish suspects merely for being inarticulate. Numerous suspects have lost Miranda rights when lower courts have followed this rule simply because, under the pressure of police custody, they haven’t been able to express their wishes clearly enough to satisfy interrogators and reviewing judges. For instance, a lower court has ruled that a suspect explaining to police that his attorney advised him to “[t]ry not to talk to anyone” is not a clear enough invocation. A 19-year-old weeping and begging to speak to his mother also doesn’t clearly express a wish to remain silent. Same with a burglary suspect saying “I got nothing more to say to you. I’m done. This is over.” As Justice Sotomayor wrote in her Thompkins dissent, “Today’s decision turns Miranda upside down.” It does so by reversing Chief Justice Warren’s requirement that the state bear the burden of showing that suspects intelligently, knowingly and voluntarily decided to submit to police questioning.
The List Grows Longer
As awareness about wrongful convictions increases, the list of similar exoneration cases grows longer. Just two months ago, for instance, the Innocence Project at Northwestern University exonerated Herman Williams after nearly 30 years of unjust incarceration. DNA testing proved that he didn’t commit the murder for which he’d been convicted. His innocence did nothing to prevent Detective Lucien Tessman from testifying at his trial that Williams had confessed to the crime. Tessman, we now know, spent his career as a serial perjurer whose specialty was lying about the goings-on in his interrogation room. Not only was his testimony about Williams’ confession false (Williams denies confessing at all), two other Illinois inmates convicted of murder due to Tessman’s lies have been exonerated.
The Fifth Amendment guarantees that the state can’t compel us to bear witness against ourselves. Chief Justice Warren issued a formula in Miranda that, if allowed, might have been successful at preserving that right. Until the Supreme Court renews its interest in solving the ancient, persistent problem of false confessions, and puts teeth back into Miranda, miscarriages of justice like those mentioned above are bound to regularly occur.