« PrécédentContinuer »
Statement of Former Prosecutors, Law Enforcement Officers, and Justice
Department Officials Dear Member of Congress:
The undersigned individuals are current and former prosecutors, law enforcement officers, and Justice Department officials who have served at the state and federal levels. Some of us support capital punishment and others of us oppose it. But we are united in our support for the federal Innocence Protection Act 2001 (S 486/HR 912).
Capital cases present unique challenges to our judicial system. The stakes are higher than in other criminal trials and the legal issues are often more complex. When the government seeks a death sentence, it must afford the defendant every procedural safeguard to assure the reliability of the fact-finding process. As prosecutors, we feel a special obligation to ensure that the capital punishment system is fair and accurate.
The Innocence Protection Act seeks to improve the administration of justice by ensuring the availability of post-conviction DNA testing in appropriate cases, and would establish standards for the appointment of capital defense attorneys. The interests of prosecutors are served if defendants have access to evidence that may establish innocence, even after conviction, and if they are represented by competent lawyers.
For these reasons, we are pleased to endorse the Innocence Protection Act. Please feel free to contact any of us to discuss this matter.
Mr. William G. Broaddus, Esq.
COLUMBIA UNIVERSITY LAW SCHOOL
July 2, 2001
Tashington, D.C. 20510
Re: Hearings on the Innocence Protection Act, June 27, 2001,
I understand that Alabama Attorney General Bill Pryor has inserted into the record of your committee's June 27, 2001 hearing on the Innocence Protection Act an October 27, 2000 speech at a bar meeting. General Pryor's speech refers to a study colleagues and I conducted last summer of rates of serious error in capital cases in Alabama and elsewhere in the United States. See James S. Liebman, Jeffrey Fagan & Valerie West, A Broken System: Error Rates in Capital Cases at http://www.law.columbia.edu/instructionalservices / liebman/liebman final.pdf (June 2000) [hereinafter, A Broken System), reprinted in part in James S. Liebman, et al., Capital Attrition: Error Rates in Capital Cases, 1973-1995, 78 TEX. L. REV. 1839 (2000) (hereinafter Capital Attrition).
General Pryor's speech contains inaccurate statements about our study and a faulty analysis of his own data on Alabama cases. I therefore respectfully request that you include this letter in the record of the hearing, immediately following General Pryor's speech, to set matters straight. I understand that the hearing record remains open for this purpose until July 5, 2001.
General Pryor's October 27 speech prompts the following nine responses (among others that would require more extended analysis than is appropriate here):
1. In our study, we showed that state and federal courts found serious error in, and reversed, 77 percent of the Alabama capital verdicts that were imposed and finally reviewed between 1973 and 1995. A Broken System, supra, at A-9. General Pryor implies that most of the errors we identified in Alabama occurred in the early part of our 1973–1995 study period, and that Alabama's rate of serious error in capital cases improved after that. In fact, the number of serious errors in Alabama capital verdicts discovered by state and federal courts during the 23-year study period held fairly steady, at about seven per year, throughout the entire study period-the beginning as well as the end. Alabama errors were not, as General Pryor suggests, front-loaded to the early part of the period.
2. General Pryor also claims that many of the serious errors found in Alabama cases were due to Alabama's refusal to permit jurors to consider whether the defendant was guilty of an offense less than capital murder, thus giving jurors the Hobson's choice of either convicting the defendant capitally or acquitting him of homicide altogether. The United States Supreme Court ruled that practice unconstitutional in Beck v. Alabama in 1980. Why General Pryor believes the frequency of Beck error in Alabama cases is a point in Alabama's favor is unclear. Beck errors are serious, because (as the Supreme Court ruled) they call into question the accuracy and integrity of the jury's decision that the defendant was guilty of a capital crime. In any event, for the following reasons, it is inaccurate for General Pryor to suggest that Beck error was the main reason that 77 percent of the Alabama capital verdicts that were finally reviewed during our study period were overturned by the courts:
(A) Most (87%) of the Alabama reversals occurred at the direct appeal stage, where duly elected members of either the Alabama Court of Criminal Appeals or the Alabama Supreme Court overturned capital verdicts because of errors of state or federal law. The single most common basis for reversal at that stage was not Beck violations (as serious as they are) but, rather, unlawful practices designed to keep African-American citizens from serving as jurors in capital cases.
(B) Over half of the remaining Alabama reversals occurred at the state post-conviction stage, following rulings by either elected Alabama trial judges or, again, by elected members of the Alabama Court of Criminal Appeals and the Alabama Supreme Court. At that stage, 67% of the errors were of three types: incompetent lawyenng, prosecutorial suppression of evidence or other prosecutorial misconduct, and jury bias. See A Broken System, supra, at C-6.
(C) The most common source of error at the third and final (federal habeas corpus) review stage was, again, incompetent lawyering.
As one would expect of the judges elected by the citizens of Alabama or appointed by the President to uphold the law, their reasons for overturning more than threefourths of the state's fully reviewed death sentences during the 23-year study period were serious and a cause for concern-not only about each of the cases in which error was found, but also about the reliability of the capital system as a whole.
3. General Pryor's speech claims that our study "does not define 'error rate' or “the basis for qualifying something as 'error.'” General Pryor did not read our study. (From the moment we issue the report last June, it has been publicly available for free to all members of the public on a number of web sites, including: http:// www.law.columbia.edu/instructionalser-vices/liebman/liebman final.pdf). The study extensively defines both “error rate” and “the basis for something qualifying as 'error.'” See A Broken System, supra, at 5–6, 25–27 & nn.33, 36, 38, 40, 42; Appendices C and D. As the study explains:
(A) Our definition of error is the courts' own definition-we made no subjective judgments of our own, and instead relied entirely on the courts' own judgments.
(B) Error defined by the courts as reversible (the only kind we counted) is serious error, because it requires a finding by a full set of courts that the defect in the case that required the state or federal courts to overturn a capital verdict and order a retrial of guilt-innocence, sentence, or both was “non-harmlesss," "actually prejudicial,” or “inherently prejudicial” as the United States and Alabama Supreme Courts have carefully defined those legal standards.
4. General Pryor suggests that he called us to ask us to define these two phrases. We never received a telephone call, e-mail, letter or other inquiry of any sort from General Pryor or any member of his staff. Had he called, we would have happily supplied him with a copy of the Report and directed him to the definitions of "serious error" and "error rates” that are a prominent part of the text and supporting materials and are cited above.
5. General Pryor says we “cannot supply you with a list of the names of all of the cases (w]e considered.” This is inaccurate. We have case lists for all 34 states in our study. Those lists will be posted on a publicly accessible web site maintained by the University of Michigan when we complete our study late this year or early next year. In the meantime, we have made the lists available to requesters, including the press, pursuant to a data-sharing agreement drafted by the General Counsel of Columbia University.
6. General Pryor says we “cannot give you a baseline of non-capital cases with which to compare his error rate.” General Pryor did not read our report. The baseline to which he refers was published in the Texas Law Review (“Capital Attrition ") at p. 1854, with sources provided in footnote 49.
7. In his speech, General Pryor says we cannot defend our “conclusion that the high error rate [w]e found means that there is a risk of an innocent person being executed .. We would like to directly study the number of innocent individuals executed in the United States in the modern era, but no such study is possible because state attorneys in Virginia and elsewhere have consistently refused to disclose the information in their files (including DNA samples) that are indispensable to any such inquiry. See, e.g., Frank Green, DNA Tests Not Likely after an Execution: Va. Opposing Third Request of its kind, Richmond Times-Dispatch, March 26, 2001. Like other Americans, therefore, we are forced to rely on evidence of risk, because the evidence of what actually happened is not available. The evidence of risk is substantial, in Alabama as elsewhere. When jurors are forbidden to convict defendants of the crime they actually committed and instead are required either to convict them of a more serious crime, of which they are innocent, or release them altogether-this being the Beck error that the Attorney General acknowledges occurred with some frequency in Alabama-that creates an obvious risk that people who did not commit a capital crime will be executed. The same is true of defendants denied a jury of their peers because African-Americans were excluded, defendants represented by incompetent lawyers, and defendants prosecuted by officials who withheld evidence of innocence. Yet, these are the very kinds of errors that were most common in Alabama, as elsewhere, during our study period. Such error puts a difficult burden on appellate courts to catch and correct all the problems. When there is so much error (e.g., in 77 percent of the fully reviewed cases in Alabama), and when the error is so serious, there is a risk-a real risk-that some of it will slip through the inspection process and never be caught. If American Airlines or U.S. Airlines had a 77 percent rate of equipment failure, we would undoubtedly conclude that the risk of innocent death is far too high-no matter how good those airlines' inspection procedures were for catching problems. The same is true here.
8. General Pryor reports his own data about all Alabama capital cases from the 1970s until October of last year. His figures confirm our findings. He reports 281 death sentences, of which 180 have not reached a final outcome in the courts, and 10 more were ended prior to a final outcome when the prisoners died in prison while the review process was continuing. Of the 91 remaining death verdicts, which were finally determined, 63 were reversed by the courts, 4 were abandoned by prosecutors as the result of settlements after errors were challenged in court, and only the remaining 124 cases were cleared by the courts. The reversal rate revealed by General Pryor's own figures thus is 67 (63 + 4) overturned out of 91 fully reviewed, or 74 percent. We found a 77 percent reversal rate for the period through 1995. General Pryor has carried the study forward through the first half of 2000, and found that the reversal rate remains about three-quarters, with only a negligible improvement in the most recent years.
9. General Pryor reports a smaller figure as the reversal rate. He gets his figure by using the 281 death verdicts imposed in Alabama, not the 91 verdicts that were actually reviewed by the courts, as his base. That is not an appropriate way to calculate an error rate because it inaccurately assumes that 100 percent of the thus far unreviewed cases (A) have been reviewed, and (B) have all been found to be without error. When 74 percent of the reviewed cases were found to contain reversible error, it is not appropriate to assume (as General Pryor does) that none of the unreviewed cases contain such errors. To use another example, if an automobile plant manufactured 281 vehicles, but only subjected 91 of them to inspections, and if 67 of the inspected vehicles were found too flawed to go to market and had to be sent back for retooling or scrap, we would not permit the plant to report a 124% error rate (as bad as that rate would be) by dividing 281 by 67. Instead, we would demand to know how many of the inspected cars were found to have serious problems-meaning 91 divided by 67, or 74 percent. The same applies here.
Thank you very much for the opportunity to insert these comments in the record of the July 27 hearing.
JAMES S. LIEBMAN Simon H. Rifkind Professor of Law
Columbia Law School
LINDQUIST & VENNUM P.L.L.P.
June 26, 2001
Dear Senator Leahy:
I have been asked to share with you and the Judiciary Committee some experience that I have had representing an innocent person who spent thirteen years on death row in the State of Louisiana. On January 2, 2001, Albert Burrell walked out of the Louisiana State Penitentiary at Angola a free' man. Unfortunately, Mr. Burrell is unable to tell you his own story using his own words. He is mentally retarded, schizophrenic and most likely suffered a serious brain injury as a child. He is also illiterate. Given his disabilities, he most likely does not know and is probably incapable of understanding the various ways in which the system failed to protect him.
In the fall of 1987, Mr. Burrell was convicted for murdering two elderly residents of Northern Louisiana over the Labor Day holiday weekend in 1986. The day after the jury determined his guilt, they unanimously agreed that he should die for his crimes. Mr. Burrell was then sent to death row at Angola. The conviction and death sentence was unanimously affirmed by the Louisiana Supreme Court. Similarly, two motions for a new trial were presented to the trial court but were rejected.
My law firm agreed to represent Mr. Burrell on a pro bono basis in early 1992. Over the years, as we were slowly granted access to the State's files on the case, our investigation uncovered substantial misconduct by law enforcement officers and prosecutors in securing the convictions of Mr. Burrell and his co-defendant, Michael Graham. Police investigation reports and witness statements that contradicted both trial testimony and the State's theory of the case had been suppressed. The State also withheld information concerning plea agreements with a jailhouse snitch that, at least with respect to Mr. Burrell, the State knew was lying. The significant government misconduct involved in Mr. Burrell's case most likely could fill a separate hearing of your committee. Nevertheless, in the summer of 1996, a warrant for Mr. Burrell's execution was issued. When we finally obtained a stay, Mr. Burrell was only seventeen days from execution.
As if misconduct by the State were not enough, a review of the record in Mr. Burrell's case also revealed a shocking incompetence by his counsel. Mr. Burrell was represented at trial by two young lawyers. The lead attorney, Keith Mullins, had been out of law school less than four years. His associate, Roderick Gibson, had been in practice less than two. Neither lawyer had ever tried a capital murder case; in fact Mr. Mullins had only handled a relatively few felony cases of any kind in his
As an initial matter, the lawyers were woefully unprepared when they began the trial. They did not investigate Mr. Burrell's significant mental health history-Mr. Mullins told me many years later when I interviewed him that he did not realize that Mr. Burrell had mental health problems. Instead, he simply thought Mr. Burrell was “a little slow.” They did not investigate the long standing child custody dispute between Mr. Burrell and his ex-wife, a critical prosecution witness who admitted to me that she lied about Mr. Burrell in order to regain custody of her son away Mr. Burrell, her ex-husband. They did not investigate a civil settlement that Mr. Burrell received from his former in-laws for injuries he received during an assault by his former father-in-law. The source of this money would have fully explained why Mr. Burrell supposedly had more money than his ex-wife had ever known him to have at the time of the murders. The State had advanced a theory that Mr. Burrell got the money from the victims after he killed them.
The defense team also did absolutely nothing to prepare for the penalty phase of the case. When it came time for that part of the trial, the defense called only one witness, Mr. Burrell, and then elicited no useful or relevant testimony. No effort was made to present any mitigating facts before the jury. Any lawyer who were to read the very short transcript of the penalty phase of that trial (it's less than 20 pages) would be ashamed for the profession.
The defense provided by these lawyers was worse than ineffective; in many instances, the defense lawyers blundered so badly that they themselves elicited damaging evidence against their own client, evidence that in at least one instance could not have been presented to the jury. For example, during Mr. Mullins' cross examination of law enforcement officers, he questioned why they had never asked Mr.