PRESS RELEASE

On Monday, February 11, another report is scheduled to be released by opponents of capital punishment, claiming to show that the system of capital trials is "broken" because of the large number of verdicts reversed on appeal. This study is a follow-up to a study released June 12, 2000, that received widespread criticism as not supporting its conclusions, stating its data in misleading ways, and, in some respects, simply dishonest. (See references at end.) That report is often called the "Liebman Report" after its lead author, Columbia Law School Professor James Liebman, a long-time opponent of capital punishment.

The fact that a large percentage of capital verdicts are overturned is not news. The controversy is, and has been for many years, whether that number reflects problems in the system for trying capital cases, as the anti-death-penalty group contends, or whether it constitutes obstruction of valid, deserved sentences, as death penalty supporters have long contended.

Although the full report is not available at this time, preliminary indications are that the follow-up report contains the same flaws as the first report. Below is a guide to those flaws, with a discussion based on a partial draft of the new report focused on whether they have been corrected. Since the report is being released in a manner calculated to hit the newspapers before the full report is available for analysis and critical review, these necessarily tentative comments are offered to provide some semblance of balance to the initial news stories. We will prepare a follow-up commentary after the Liebman group makes its full report available and we have an opportunity to analyze it.

Ignoring Erroneous Reversals

One of the largest, on-going problems in capital litigation is the erroneous overturning of valid sentences by courts hostile to capital punishment. The U.S. Court of Appeals for the Ninth Circuit, with jurisdiction over nine western states, has been particularly notorious in this regard. A 1995 study by the Criminal Justice Legal Foundation looked at cases where that court overturned sentences based on disagreement with state courts on an issue of capital sentencing law, and where the U.S. Supreme Court subsequently resolved the disagreement. On all but one of a dozen issues, the state court decision upholding the sentence was correct, and the federal court decision finding "error" was itself erroneous. For example, the Ninth Circuit overturned the death sentence of organized crime "hit man" John Harvey Adamson for the car-bomb death of Arizona reporter Don Bolles on the ground that Arizona's death penalty for "especially cruel" murders was too vague. Two years later, in a different case, the Supreme Court held that the Arizona law was valid. Unfortunately, there are far more cases than the Supreme Court can review, so erroneous reversals such as Adamson very often stand.

The first Liebman report simply ignored this problem. It counted as "serious error" every finding causing reversal of a conviction or sentence. There is no indication in the available materials for the second report that any attempt has been made to distinguish valid from erroneous reversals. On the contrary, the new report looks at the low rate of reversals in California state courts and the large number of these cases subsequently overturned by the Ninth Circuit, and concludes the federal court is making up for "lax" review by the state court. The Ninth Circuit is the one court in the nation most often reversed by the Supreme Court, and the obvious alternative explanation is that the Ninth is wrongly overturning correct judgments. It does not appear that the new report even considers that possibility.

Constantly Changing Rules

In legal jargon, a judgment may be deemed in "error" if it is contrary to the rules as they exist at the time of the appeal, even if it was perfectly valid under the rules in effect at the time it was rendered. For over 25 years, the Supreme Court and other courts have continuously tinkered with the rules, and all of the changes apply retroactively to all cases still pending on the first round of appeals. Justice Scalia aptly called this the high court's "annually improvised" jurisprudence. Here are a couple of examples. A court instructs a jury in accordance with a statute the Supreme Court has just upheld as valid, and nine years later that instruction is declared constitutional "error." Another court uses a standard instruction and verdict form telling jurors they must deliberate and agree on the circumstances to be weighed in reaching their verdict, in complete accord with the long-standing American tradition of jury decision-making. Years later, out of blue sky, that instruction and form are declared invalid, and all of the cases that used it (which may be all the capital cases in the state) are suddenly in "error" and must be retried.

These "errors" do not indicate anything at all wrong in the trial court, and their existence should not undermine public confidence in capital trials in the slightest. Yet the vast majority of them are included in the study's definition of "serious error." The new study indicates that it excludes the cases where the Supreme Court has declared a state's entire system unconstitutional, but that has not happened since the mid-1970's. Far more common is a decision throwing out a standard instruction, form, or practice that had previously been considered perfectly valid.

The new report decries the waste and delay that are caused when so many judgments are reversed, and supporters of capital punishment wholeheartedly agree. But nothing we can do at the trial level will prevent reversals of this type. The only answer is for the reviewing courts, and especially the Supreme Court, to stop inventing new restrictions. Whatever the intrinsic merit of these rules may be, the turmoil of the change exacts an enormous cost.

Blurring Guilt and Punishment

Several commentators criticized the first report for glossing over the distinction between the determination of guilt of murder and the determination that the particular murderer ought to be sentenced to death. (See the Wilson article and the Latzer and Cauthen articles, cited below.) Most people would agree that the execution, or for that matter the imprisonment, of an innocent person is of far greater concern than the execution of any person who is actually guilty of premeditated murder. The question of greatest concern is the degree to which the system risks executing a person who neither killed the victim nor was a party to the killing. The "abuse excuse," "Twinkie defenses," rules excluding valid evidence because of how it was obtained, and compliance with the Supreme Court's Byzantine code of sentencing procedure are all matters of much lesser moment.

The first report told us very little along these lines. It does not appear that the new report will add much. For the most part, it lumps guilt and sentence reversals together. The report does indicate that 9% of the cases sent back for retrial of guilt verdict ended in acquittals. That is, these cases are retried, typically a decade or more after the fact, when memories have faded and witnesses may no longer be available. In some cases evidence used the first time is suppressed for reasons unrelated to its reliability, such as the Miranda rule. In a small percentage of these cases, the jury decides that guilt has not been proven to the exacting standard of "beyond a reasonable doubt." Our trial system is intentionally stacked in the defendant's favor in many ways, including the burden of proof and the fact that the prosecution cannot appeal trial errors. Many guilty people are acquitted as a result, and a handful of acquittals among the retrials would be expected even if 100% were actually guilty. The fact that the acquittal rate on retrial is so low serves to reinforce confidence in the system, not undermine it.

Intentionally Skewed Sampling

A major theme of the first report was to convince the public that incompetent lawyers for capital defendants and suppression of exculpatory evidence were the main problems. To this end, Columbia Law School put out a press release announcing the study with this statement:

"The study found that the errors that lead courts to overturn capital sentences are not mere technicalities. The three most common errors are: (1) egregiously incompetent defense lawyers (37%); (2) prosecutorial misconduct, often the suppression of evidence of innocence (19%); and (3) faulty instructions to the jurors (20%). Combined these three constitute 76% of all error in capital punishment proceedings."

This statement was a patent falsehood. Those percentages are not percentages of the total, but only of a narrow segment of cases, those overturned in "state post-conviction" review. That is the stage of the process particularly geared to claims of ineffective assistance and nondisclosure of evidence. Analogously, if a researcher stations an observer in the tire shop of an auto center, he will observe that most of the cars repaired there have tire problems. That observation, while true, means nothing.

The new study continues the effort to exaggerate the number of cases of defense lawyers deemed ineffective by the Monday-morning quarterbacks and of prosecutors who failed, often inadvertently, to turn over a piece of evidence that in hindsight might have made a difference. The mechanism, again, is the skewed sample. This time the sample is extended to include federal as well as state habeas corpus review, but the result is largely the same. Cases on direct appeal are still excluded from the analysis of the reasons for reversal, even though that is where 80% of the reversals occur. Direct appeal is, not coincidentally, also where reversals for "mere technicalities" most often occur. It is not hard to get the results one wants if one can exclude the 80% of the cases where the inconveniently contrary data points are likely to be found.

The excuse offered by the study summary is that state and federal habeas corpus cases were selected because that is where data was "available." That assertion is not credible. Direct appeal is far and away the easiest segment of the process to track. Capital cases in most states are appealed directly to the state's highest court, which usually publishes all of its opinions. Cases on state or federal habeas corpus, by contrast, tend to be dispersed among multiple courts, and the petitions are far more likely to be disposed of without published opinions, or often without any opinions at all. The selection of the skewed sample has the distinct odor of intentional distortion.

References:

James Q. Wilson, What Mistakes in Death Penalty Cases?, New York Times, July 10, 2000, Section A, page 19, column 1.

Joseph L. Hoffman, Violence and the Truth, Indiana Law Journal, volume 76, number 4, page 939 (Fall 2001).

Barry Latzer and James N.G. Cauthen, Capital Appeals Revisited, Judicature, volume 84, number 2, page 64 (Sept.-Oct. 2000).

Barry Latzer and James N.G. Cauthen, The Meaning of Capital Appeals: A Rejoiner to Liebman, Fagan, and West, Judicature, volume 84, number 3, page 142 (Nov.-Dec. 2000).

Kent S. Scheidegger, Death Penalty Study Is Full of Unjustified Assumptions, Los Angeles Daily Journal, June 30, 2000, page 6, col. 1.