Dobie Gillis Williams was picked up in a police roundup after a woman was stabbed to death in her bathroom in rural Louisiana in 1987. With an IQ of 65, and a lawyer who was subsequently disbarred for incompetence, Williams was tried for murder, convicted and sentenced to death. He twice came within hours of execution, only to have the process stayed, first by the U.S. Supreme Court and a second time by the governor.
A federal judge finally overturned Williams' sentence, citing two errors: Williams had confessed only after a police officer promised Williams he wouldn't get the death penalty, and his attorney never offered any evidence of mitigating circumstances during the sentencing phase of his trial -- evidence that should have included Williams' mental deficit and a childhood of abuse and neglect.
But Williams was executed anyway in 1999, when the notoriously pro-death-penalty 5th Circuit Court of Appeals (which serves Louisiana, Mississippi and Texas) overturned the district court judge's decision, ruling that the revelation of errors came too late under the 1996 Effective Death Penalty Act (EDPA). Williams' appellate attorney, Nick Trentacosta of the Louisiana Center for Equal Justice, maintains that his client's execution was in error. The fact that mitigating evidence hadn't been presented to the jury wasn't in dispute -- just the timing of when that issue had first been raised on appeal.
Williams, whose case might seem unique in its disastrous outcome, was hardly unusual. Nobody has yet come up with an irrefutable example of an innocent person having been executed since states began reinstating the death penalty in 1976, but even death penalty advocates concede that sentencing errors in death penalty cases have led to the execution of many criminals who, by law, should not have received the ultimate sanction.
"There are definitely plenty of sentencing errors where those who die haven't deserved to die," says Robert Blecker, a self-described "staunch advocate" of the death penalty who teaches criminal law at New York Law School. Blecker, who believes that errors work both ways -- with people who should, in his view, be executed getting off -- adds that the main cause of wrongful executions is inadequate counsel and the failure to present mitigating evidence to the jury.
And, he says, such sentencing errors "tend to happen most often in cases where those who die don't really deserve to die." In the most heinous murders, he claims, sentencing errors are rare.
Appellate courts have found sentencing errors in more than 40 percent of all capital cases since 1973 -- when the Supreme Court ruled the existing death penalty statutes unconstitutional -- but there are no figures to indicate how many inmates were executed or remain on death row due to sentencing errors. Steven Hawkins, a staff attorney with the National Coalition to Abolish the Death Penalty, estimates that "at least 40 to 80" of the 800 people executed since 1977 "were probably victims of obvious error, such as failure to have mitigating circumstances raised at trial." An equal number, he adds, were "simply victims of the U.S. Supreme Court's failure to make various death penalty rulings retroactive" for all condemned inmates.
That last condition -- essentially, the stubborn upholding of fatal errors by the courts -- is remarkable, given the stakes. The decision to apply rulings exclusively to future cases, largely to avoid disrupting legal precedents, is typical of the U.S. Supreme Court. Most often, it is barely noted. But in death penalty cases the procedure has enormous impact. When a sentencing error has been exposed, often as an unconstitutional element of existing law, but not applied retroactively, it means that those on death row as a result of the same unconstitutional law will not be spared -- unless their case was specifically part of the appeal that brought the ruling.
Just last year, for instance, the Supreme Court ruled it was unconstitutional for death penalties to be determined by trial judges. Rather, the court held, that decision belongs to juries alone. Yet a number of condemned men -- in states such as Alabama, Arizona, Colorado, Delaware, Florida, Idaho, Indiana, Montana and Nebraska -- have subsequently been executed after being sentenced to death by judges -- sometimes, as in the case of Alabama, Delaware, Florida and Indiana, by judges who actually overruled juries that had recommended a lesser sentence. More prisoners, with older cases, are still being sent to their unconstitutional death. In fact, this is true for many, if not most, of the Supreme Court's more far-reaching death penalty decisions.
The same holds for lower court decisions. In December 2001, a federal District Court in Pennsylvania overturned the death sentence of death-row prisoner Mumia Abu-Jamal, on the grounds that his jury in 1982 had been provided with a sentencing form so confusing that it appeared to require all 12 jurors to agree to a mitigating circumstance before any one of them could consider it in deciding whether Abu-Jamal deserved the death penalty or life without parole.
Abu-Jamal, a journalist and former Black Panther convicted of killing a white policeman, came within days of being executed back in 1995 because of this error, but he was spared when his appeal hearing ran past the execution date, voiding it. In fact, Pennsylvania, like most other states with death penalty statutes, says that if any juror finds a mitigating circumstance, that juror may consider that factor in deciding between life or death. (Since a death sentence must be decided unanimously by a jury, just one juror voting against death results in a life sentence instead of execution.) Yet despite the reality that many of the 244 other people on Pennsylvania's death row were sentenced by juries using the same confusing form, no blanket order revoking those sentences was issued following the ruling in Abu-Jamal's case.
Jim Liebman, a professor of law at Columbia University, is the lead author of two recent studies of death penalty errors, which surveyed virtually all the death penalty cases in the country since the Supreme Court reinstated capital punishment in the mid-1970s. He points out that the highest rates of sentencing error are consistently found in states such as Arizona, Nevada or Pennsylvania, all of which have very "broad" death penalty statutes -- that is, statutes that allow for capital punishment for a wide variety of crimes -- while a state like Colorado, which has a very narrowly applied death penalty, has far fewer errors.
In reviewing all 5,760 death sentences issued between 1973 and 1995, Liebman says he found that 41 percent -- about 2,360 cases -- had been tossed out on appeal because of "sentencing errors." Among those, more than 300 were overturned on initial appeal by a state court on the grounds that the so-called aggravating circumstance (a finding required for imposition of the death penalty) in fact did not exist. A third were overturned at the post-conviction phase or by a habeas appeal on grounds of "ineffective counsel" (most often for failure to introduce mitigating evidence at a sentencing hearing). Another 19 percent were overturned because of the suppression of exculpatory evidence by prosecutors.
Subsequent to their sentences or convictions being overturned, 82 percent of capital defendants have received lesser sentences -- a figure that leads Liebman to conclude that "they were oversentenced in the first place." Liebman notes that sentencing errors are also likely to be common among the 59 percent of capital cases in which the convictions were not reversed on appeal.
While public debate and media attention about the death penalty have focused on wrongful convictions, "sentencing errors go to the heart of the problem with the death penalty," Liebman argues. "The idea of the death penalty is that it is supposed to be applied to the worst of the worst offenders, but this is not what is happening." He adds that "hundreds of cases of sentencing error" have gone uncorrected, including many that have already led to executions. Agreeing with Blecker, he adds, "The marginal cases are where most of the errors turn up."
Advocates of capital punishment, for the most part, express little concern about sentencing errors. "Just about anyone would agree that the execution, or for that matter just the incarceration of an innocent person, is of far greater concern than the execution of a person who is actually guilty of premeditated murder," says Kent Scheiddegger, of the Criminal Justice Legal Foundation, a pro-capital-punishment organization based in California. "I don't think that the execution of a person who has committed premeditated murder is an injustice. The worst it can be is an uneven dispension of mercy."
Others see the matter differently. "Even if it is someone who is a killer, we're still talking life and death," Liebman says. "I don't think we can be so callous as to say, 'They're killers, so who cares if they live or die' -- especially since you see people with resources getting their sentences overturned and others without resources going on to execution."
Besides, not everyone sent to death row is guilty of premeditated murder. Take Phillip Tompkins, a borderline mentally retarded black man convicted of killing a woman in Texas after deliberately colliding with her car and then taking her to a secluded spot to rob her. According to prosecutors, Tompkins tied the woman, Mary Berry, to a tree, stuffed part of a bedsheet in her mouth, and then left to use her ATM card and steal $1,000. An all-white jury found that Tompkins had murdered her, but it was never clear whether he had intended to do so, particularly given his low IQ.
Manny Babbitt suffered the same fate. A decorated Vietnam War veteran (he was awarded his Purple Heart on San Quentin's death row only weeks before his death), Babbitt was executed in May 1999 for the death of Leah Schendel, a 78-year-old woman who died of a heart attack in 1980 during Babbitt's attempt to burglarize her home. No claim of premeditated murder was made by the state. Rather, it became a death penalty case because of a California law making death caused in the process of a burglary a capital offense.
Babbitt's state-appointed trial attorney -- who was later disbarred for financial improprieties -- failed to introduce mitigating evidence of his client's prior confinement in a mental hospital, or of his obvious symptoms of post-traumatic stress disorder from his Vietnam experience -- Babbitt had been observed reenacting battle scenes as a homeless person in Providence, R.I., prior to the burglary.
Two jurors in the Babbitt case subsequently testified that had they known of his wartime experiences and mental history, they would not have voted for death. His subsequent execution, which followed a rejection by Gov. Gray Davis of his clemency appeal (which had been endorsed by hundreds of Vietnam vets), also illustrates how capital-sentencing error, like murder itself, casts a wide net of suffering.
"When my Manny was sentenced to death, at first I thought of killing myself," recalls his older brother Bill, 60, who personally turned in his brother after word of Schendel's death, on a promise by the police that his brother would not receive a death sentence. "I turned Manny in because I felt sympathy for that family, and I felt very guilty about it when they went ahead and sought the death penalty."
Bill Babbitt, now active with the group Murder Family Victims for Reconciliation, says that his brother's 17 years on death row, and his ultimate execution, were "devastating" to his family, which included himself, a mother, two children and several grandchildren. While saying he feels great sympathy for the victim's family, he asks, "What kind of closure did they have? They could have had closure in 1982 if Manny had been sentenced to life in prison. The prosecutors told them that when Manny was dead, they'd feel better. Instead, the state kept that wound open almost 20 years. That's closure?"
Bill Babbitt watched, along with several of Schendel's relatives, as his brother was killed by injection. "Her granddaughter couldn't watch," he recalls. "She looked at the floor. So did the prosecutor. But I watched the whole thing. My brother had told us he was not afraid to die, and that we should not show any malice toward the victim's family. He said he would keep his eyes closed and meditate on God and Jesus, but that he wanted a smiling face watching him. So I told him I'd be there."
Liebman says that since sentencing errors are so often the result of incompetent or inadequate defense attorneys, they are often red flags for probable errors in the guilt phase of the trial -- suggesting that some of those with sentencing errors may well have been wrongfully convicted too. "It's just that sentencing errors are much easier to prove," he explains.
That is because while guilt must be proven "beyond a reasonable doubt," the standards for justifying the death penalty are much lower. "In the conviction phase of a capital trial, the whole burden of proving guilt is on the prosecution, so you pretty much have to prove the errors occurred there, which is harder to do," says Liebman. "In the sentencing phase, part of the burden is on the defense -- to bring in mitigating evidence to convince a jury that death is not appropriate. If an attorney doesn't do that, it's easy to point it out on appeal."
That doesn't mean an appeals court will go along with a claim of ineffective counsel, however. Even in cases where defense attorneys slept through trials or attended while stone drunk, higher courts have upheld the subsequent convictions and death sentences of their hapless clients. But at least it is relatively easier to present evidence that no mitigating evidence was submitted than to prove that a conviction was incorrect.
Liebman adds it is also politically easier for judges in capital cases to overturn a sentence that leaves a convict behind bars for life than to overturn a conviction, which raises the possibility that the individual might end up back on the street. At the same time, he says it is clear that, since the reinstatement of the death penalty in the United States nearly three decades ago, hundreds of capital prisoners have already been executed whose sentences were in error.
That certainly was the case with Robert Coe, the only prisoner to be executed in Tennessee since the reinstatement of the death penalty in that state. Coe, executed in 2000, was a severely mentally ill individual with a grotesque history of abuse as a child -- his father raped his sister in front of him and had punished him for a childhood transgression by deliberately shooting him in the leg.
While there were serious questions about whether Coe was even guilty of raping and killing a young girl -- the police found bloody sheets in the room of a prosecution witness whom the defense argued might have been the real killer, but they "lost" the evidence before testing the blood -- what is beyond debate is that Coe's public defender, like Dobie Gillis Williams' attorney, never introduced any mitigating evidence during his sentencing hearing. The jury never heard that he had been institutionalized in Florida for severe schizophrenia, nor did jurors hear anything about Coe's history of child abuse at the hands of his father -- evidence that ought to have mitigated against a death sentence.
Attorneys brought into the case late in Coe's appeal to the federal courts tried to introduce the mitigating evidence, but the federal courts, bound by the strictures of the Effective Death Penalty Act of 1996, held that it was "too late" for such evidence to be considered. They ruled that Coe's original trial attorney and his first appellate attorneys had erred in not introducing that evidence at the state court level, and that it was thus "procedurally barred." (In an all too common Catch-22, the court also held that the seemingly outrageous error of not introducing readily available mitigating evidence was not sufficient evidence of ineffective counsel.)
As in Williams' case, a federal district judge did temporarily lift Coe's death sentence on the grounds of an erroneous jury instruction, but a three-judge panel of judges in the 6th Circuit Court of Appeals overturned that ruling on a 2-1 vote, and Coe was executed.
"The state of Tennessee decided to execute a mentally ill person. They knew he was mentally ill and that he had been an abused child, but they decided to go forward with Tennessee's first execution in 40 years," says the Rev. Joe Ingall, who came to know Coe during his years of incarceration on death row. "This case was all about the politics of death, not about justice. They needed to kill somebody, so they killed Robert."
After years of clamping down on death row inmates' right of appeal, the Supreme Court seems to have lightened up in recent months, making capital-punishment cases a little harder to win and a little easier to challenge. Besides its ruling that juries alone can vote for a death sentence, the court more recently lowered the standard of proof for death row inmates seeking to claim that their jury panels were unconstitutionally purged of minority jurors. The court also criticized the 5th Circuit Court of Appeals and lower federal courts in the district for accepting without question the rulings and procedures of state courts in death penalty cases, suggesting that "technicalities," which can be life-or-death matters in capital cases such as Coe's and Williams', while an important aspect of American jurisprudence, are not clear-cut matters of fact.
Finally, the court ruled that retarded people should not be executed, though it left unresolved how lower courts should determine what constitutes retardation.
But capital-punishment critics like Liebman, and even death penalty advocates like Blecker, argue that the problem of errors, in both the guilt and sentencing phases of capital cases, will continue to plague the system, at huge cost -- both psychological and financial -- to society and to inmates, and to the families of victims and inmates.
"Everybody loses," says David Elliott, a spokesman for the National Coalition to Abolish the Death Penalty. "Remember that people on death row have relatives, and those relatives are losing a loved one. And the victim's relatives have to live through years of trials and hearings. And then there's the prisoner," he adds. "If he's executed because of a sentencing error, that's terrible, but even if the error is eventually corrected, he has lived under the threat of death for years. We have seen these people, and afterwards many of them have something resembling post-traumatic stress disorder. They can't adjust to the reality that they aren't going to die."
Both Liebman and Blecker suggest that the quickest way to reduce these errors would be to ensure that all capital murder defendants are provided the resources to put on an adequate defense, and for legislatures and prosecutors to narrow the range of crimes that allow for a death penalty.
But mandating adequate defense counsel in capital cases is costly, and in most states, the burden of financing first-class defense counsel and appellate counsel for capital cases would fall on already strapped local governments. Raising taxes is always a hard sell. Raising them in order to help capital defendants beat the rap, or to help convicted killers contest their convictions or their sentences is an even harder sell. Yet critics of the system say something has to be done.
"When you have a death penalty system, and you start diluting it by executing people who don't deserve to die while not executing others who do, it means the system is breaking down," says Liebman. "What's needed is more honesty about what's happening. We have to have a commitment to get rid of all the error."
This story has been corrected since it was originally published.