British comedian Sacha Baron Cohen, appearing in character as Admiral General Haffaz Aladeen to promote his new film, The Dictator, has been working in lines about the double standards of American human rights assessments. "What in Wadiya you call genocide," he says, referring to the dictator's fictitious Arab homeland, "in Texas you call the justice system." Texas criminal justice may not amount to genocide, but it does misfire with alarming frequency, and claims innocent lives in the process.
Since 1982, when Texas resumed judicial killings following a moratorium during the Sixties, the state has executed 482 people — four times the number of the next most aggressive state. Moreover, amid the rough-and-tumble of Texas's right-leaning political culture, a candidate's willingness to authorize large numbers of judicial killings and his distaste for clemency and pardon reviews has become a sine qua non for holding the state's governorship—which has in turn become an important launching pad for G.O.P. presidential candidates, particularly since Texas is the most populous predictably Republican state.
Now, the Columbia Human Rights Law Review has published a book-length study by Columbia law faculty and students exhaustively examining one of those executions and concluding that Texas very likely executed the wrong man. The study, titled Los Tocayos Carlos (The Carlos Look-alikes), goes beyond vindication of the executed man to identify the actual perpetrator of the crime, backed by a convincing stream of evidence.
On February 4, 1983, Texas prosecutors charged Carlos DeLuna with the knifing murder of convenience-store attendant Wanda Lopez in Corpus Christi. DeLuna was put on trial and convicted. Texas appeals courts sustained the conviction in a rapid-fire series of decisions. A last-minute appeal to Texas governor Bill Clements for clemency went nowhere after DeLuna's lawyer, Kristen Weaver, told the governor's counsel that this was a "case where guilt was clear and obvious." In December 1989, DeLuna was executed by lethal injection at the state penitentiary in Huntsville.
A team of Columbia law students led by professor James S. Liebman started probing the DeLuna case in 2003 after surveying past Texas cases for convictions in which eyewitness identifications played an essential role. Once viewed as a key indicator of a valid conviction, eyewitness identifications are increasingly being viewed by psychologists and criminal-justice scholars as a weak link in the system. DeLuna had claimed during his defense that a "look-alike Carlos" actually committed the crime — a claim that prosecutors and reporters dismissed, and that wasn't even seriously followed up on by his own counsel. Yet a Columbia investigator's visit to Corpus Christi quickly turned up astonishing evidence: a woman who knew both Carloses and who had heard a story, handed down in family accounts, in which the other Carlos bragged about how he had made his double take the fall for a homicide. The story of this other Carlos, Carlos Hernandez, quickly fell into place, as the Columbia team uncovered evidence and details corroborating DeLuna's account. Without drawing formal conclusions, the study suggests that Texas prosecuted, convicted, and executed the wrong man. The likely murderer, Carlos Hernandez, died of cirrhosis of the liver in May 1999 while he was in prison for narcotics charges.
"One of the most disturbing things about the DeLuna case is that it was not a special or notorious case," Liebman told me in an interview. "Everyone ignored it because it was so routine, because the convicted man's guilt seemed so clear, because his story seemed so weak." Liebman notes that even as DeLuna was sent to his death by a prosecutor who derided his defense as a "phantom," a second prosecutor sat silently by in the courtroom, knowing that there really was a Carlos Hernandez and that he matched DeLuna's description. This second prosecutor had investigated Hernandez three years earlier for another knifing murder of a Hispanic woman. The Columbia investigation also found that prosecutors had obtained and examined Hernandez's rap sheet—in other words, they had misled the jury about the "phantom," making a false argument that led to DeLuna's conviction and execution.
Texas police also had a reason to deflect attention from Hernandez — he was likely working for them as an informant. Indeed, the Columbia report documents a number of serious crimes in which Hernandez was implicated and inexplicably escaped charges, suggesting that his relationship with local law enforcement might have been deep and complex.
Liebman also believes ethnicity might have played a role in the miscarriage of justice, in two different ways. "Both the defendant and the victim were Latino and not important persons in the eyes of the authorities," he told me. "Lopez was victimized twice — the first time when she was knifed, and the second time when police and prosecutors undertook a shoddy investigation and pushed to convict an innocent man." Did Texas execute an innocent man, I asked? "All of us who worked on the project believe that it did."