February 14, 2002 

The link between death-sentencing rates and reversal rates was found in an analysis of more than 5,000 death sentences fully reviewed by state and federal courts from 1973 to 1995. 

The 10-year study was financed by Columbia University Law School and directed by Professor James Liebman, a capital punishment and habeas corpus scholar there. 

"We think there are a number of pressures to use the death penalty where the focus is really a bigger social problem than a particular case," Liebman said. He said that the data also revealed a number of other important factors related to high capital reversal rates, including large black populations, a high risk of homicide to whites, partisan judicial elections and poor law enforcement generally. For example: 

• Reversal rates are twice as high where murder victims are white than where murder victims are black. 

• The more frequently that state trial judges are subject to direct election and the more partisan those elections are, the higher the rate of serious capital error. 

• The lower the rate at which states imprison serious criminals, the higher their capital error rates. 

The link between death sentences and reversible errors was seen most sharply in a comparison among counties with 600 homicides or more, said Liebman. 

Ten counties with the highest death-sentencing rates had an average capital error rate of 71 percent at the first and last appeal stages. Eight of those counties sent 16 people to death row who were later exonerated. In 10 counties with the lowest death-sentencing rates, the average error rate was 41 percent, and no one sentenced to death was later found not guilty. 

AVOIDING REVERSAL

The data suggest that policymakers should consider making it difficult for prosecutors to use the death penalty in other than the "worst of the worst" cases, Liebman said. Aggravating circumstances should "substantially outweigh" mitigating circumstances, he said. For every additional aggravating factor, he said, the error rate drops 15 percentage points. 

"You can also try to carve out categories of cases at the margin and put them beyond capital eligibility, for example, cases with mental retardation or juvenile cases," he said. 

He said that policymakers should also consider a proposal made in Oklahoma to require a jury to find guilt "beyond any doubt." And life imprisonment without parole should be an alternative sentence to death, he said. Judge overrides of juries' life sentences should be abolished in those states using them. 

To reduce political pressures on judges, the study suggests that trial and appellate judges should be appointed rather than elected. If judges are elected, it says, nonpartisan elections or retention votes should be used instead of contested elections. 

Capital trial and appellate judges, appointed or elected, should have terms lengthened. And jurors, not judges, should determine the sentence in a capital case, he said. 

The study establishes what many practitioners have found over the years, said John Blume, director of the Cornell Death Penalty Project at Cornell Law School. There really is no "Texas death penalty" or "South Carolina death penalty." he said. 

"You do get these statewide patterns but when you peel it back, what you see is it's very much a county or judicial circuit phenomenon," he said. "Even in places like Texas, there are many, many counties which have never imposed a death penalty. 

"What you have, by and large, are prosecutors who are very pro-death in select places in these states who are driving this. And that raises a lot of questions that I don't claim to have answers to." 

The study is Part Two of the Columbia project. It represents an effort to understand why, as Part One (see "also of interest" item) reported in June 2000, state and federal appellate courts reversed 68 percent of more than 5,000 of the death sentences studied. 

Reforms will have to be state by state, said Liebman, and proposals are being considered in several.