Reviewing all 4,578 cases from 1973 through 1995 in which a final decision was reached, the study found that 68 percent were reversed by state or federal courts. In those cases that were retried, 82 percent of the defendants received a lesser sentence than death and 7 percent were found not guilty.

Florida's overall error rate of 73 percent was slightly higher than the national average. In the 23-year period, the state had 870 death sentences, which resulted in only 36 executions because of appeals, reversals and delays.

The study focused on 26 states with the death penalty where there have been fully completed appeals, but five of the states had three or fewer federal appeals. Florida's ranking on reversals fell near the middle.

``Our results reveal a death-penalty system collapsing under the weight of its own mistakes,'' said Professor James Liebman, who oversaw the study. ``They reveal a system that is wasteful and broken and needs to be addressed.''

The average time from sentence to either execution or final resolution was nine to 10 years nationwide and about the same for Florida, the study found. Liebman said his review showed that serious errors often are found many years after the first trial.

``This much error, and the time needed to cure it, impose terrible costs on taxpayers, victims families, the judicial system and the wrongly condemned,'' Liebman said.

The new study probably will add fuel to the growing debate over the fairness, efficacy and speed of the death-penalty process around the nation. The study took years to complete because data from each state and each level of the appeal process are kept differently, or not at all.

Thirty-eight states have capital punishment, but six -- New Mexico, South Dakota, New York, New Hampshire, Connecticut and Kansas -- have executed no one since 1976.

CONCLUSIONS VARY

While other reports have focused on high-profile exonerations in Florida, Illinois and elsewhere, the Columbia study looked at all cases that were resolved over a 23-year period. Leaders in the death-penalty debate in Florida were quick to draw different conclusions from the same numbers.

Gov. Jeb Bush has complained that the state Supreme Court is taking too long to review death cases and that delays are getting worse, with an average now of about 14 years from sentencing to execution. The high court in April struck down efforts by Bush and the Legislature to sharply reduce the time for appeals.

A Bush spokesman, Justin Sayfie, said the Columbia study shows that Floridas ``extensive appeals procedure, with adequate due process, works in reviewing cases.''

``Its important to remember that reversals dont mean innocence,'' Sayfie said. ``There may have been all sorts of legal issues that send a case back.''

'FINDING ERROR'

Columbia's Liebman also said Florida courts might be better at catching mistakes than those of some other Southern states.

The 11th U.S. Circuit Court of Appeals, which provides a final level of review, reversed only 37 percent of the Florida death sentences it reviewed in the 1973-95 period, compared with 45 percent for Alabama and 65 percent for Georgia.

``That shows that Florida courts are doing a good job, relatively speaking, in finding error,'' Liebman said.

In the past 27 years, 87 inmates around the country have been released because they were wrongfully convicted, and Florida leads the nation with 19 exonerations.

But Gerald Kogan, the former Florida chief justice who reviewed dozens of death cases on the high court, said the Columbia study reinforced his perception that the administration of the death penalty is fraught with problems.

``This is the ultimate penalty, and too many mistakes are being made,'' Kogan said. ``Reversal at our level, or federal, requires serious error.''

TIME-CONSUMING

Kogan said that while state Supreme Court review of death cases has been ``thorough and stringent,'' that requires time. He estimated that death-penalty appeals constitute 3 percent of the high courts caseload and take up half the justices time.

Kogan, who chairs a new national group, the Committee to Prevent Wrongful Executions, said the study provides data on the large number of cases ``where the death penalty is just not appropriate.''

In Florida, he said, prosecutors, defenders and judges in large urban areas have experience in dealing with a wide range of homicides, but in smaller communities, any homicide, even a crime of passion, can bring a death sentence that often gets thrown out later.

Liebman, in an interview last week, said that before execution, most Death Row inmates cases go through three levels of review -- direct appeal, another review before a states high court, and a federal habeas corpus review. ``Serious error'' was caught at one of those levels in 68 percent of the cases, he said.

REASONS SOUGHT

Liebmans Columbia Law School team is conducting a follow-up study on the reasons for reversals.

Liebman said a preliminary analysis of reversals in Florida showed that the leading causes were ``grossly incompetent defense counsel'' (28 percent), bad jury instructions by the trial judge (21 percent) and prosecutor or police misconduct (18 percent).

``I tried to approach this not from the standpoint of morality or fairness, but as a bean-counter or manager,'' said Liebman, who as a defense lawyer has argued four death-penalty cases before the U.S. Supreme Court.

``As taxpayers, can we accept this level of error, and are we catching them all?'' he asked. ``If you were an airline passenger told that serious problems were found in 68 percent of aircraft, would you be confident about getting on a plane out of the remaining 32 percent?''

Copyright 2000 Miami Herald

###