• Sheet of paper Report

USA: Less than “ironclad,” less than safe

March 26, 2011


USA: Less than ‘ironclad’, less than safe

Federal court ruling may clear way for Georgia to set new execution date for Troy Davis

27 August 2010


AI Index: AMR 51/077/2010

[W]hile the State’s case may not be ironclad, most reasonable jurors would again vote to convict Mr Davis of Officer MacPhail’s murder. A federal court simply cannot interpose itself and set aside the jury verdict in this case absent a truly persuasive showing of innocence

– Judge William T. Moore Jr., US District Court for the Southern District of Georgia, 24 August 2010



If a state’s case against a condemned prisoner is not “ironclad”, should not that fact trouble those pursuing his execution? This is a question that should be asked of the authorities in Georgia, USA, following a federal judge’s ruling that, if upheld, will clear the way for the state to kill Troy Davis in its execution chamber.

US District Court Judge William Moore answered a different question in his ruling issued on 24 August 2010. He addressed not whether the state could demonstrate a watertight case against Troy Davis, but whether Troy Davis could show “by clear and convincing evidence that no reasonable juror would have convicted him in the light of the new evidence” that has emerged since his 1991 trial for the murder in 1989 of Savannah police officer Mark Allen MacPhail.

Under this “extraordinarily high” standard, Judge Moore wrote, “Mr Davis is not innocent”.

Elsewhere in his ruling, Judge Moore displays less certainty in the state’s case than this bare conclusion would otherwise suggest. He acknowledges that the new evidence presented by Troy Davis casts “some additional” doubt on his conviction, albeit not enough to justify reversal of the jury’s verdict. And in the final footnote to his 174-page ruling, he writes that while the state’s case against Troy Davis may not be “ironclad”, his own review of the evidence had led him to conclude that “most reasonable jurors” would vote to convict Troy Davis. Evidently, he believes that there is doubt enough for somereasonable jurors to vote to acquit Troy Davis on the current evidence.

The State of Georgia had urged Judge Moore to place an “extraordinarily high” burden of proof upon Troy Davis. In electing to do so, the Judge explained that choosing a standard that was not only this burdensome, but also “crafted from the perspective of a reasonable juror”, “comports with the high level of respect society has for jury verdicts rendered subsequent to an uncorrupted process, while acknowledging that even the best efforts of society may occasionally yield results that later prove clearly incorrect”.

The 1991 jury had found Troy Davis guilty “beyond a reasonable doubt,” Judge Moore noted, “but not to a mathematical certainty”. The fact that the beyond-a-reasonable-doubt standard cannot guarantee error-free verdicts is highlighted by the cases of the more than 130 people who have been released from death rows on grounds of innocence in the USA since 1976. In each case, the defendant had been found guilty beyond a reasonable doubt. Errors, then, have occurred more than “occasionally”. For every 10 executions in the USA since 1976, one death row inmate has been exonerated. How many innocent prisoners have been executed during this period is unknown.

The Troy Davis case is one in which many of the witnesses who testified against the defendant have since retracted or contradicted their trial testimony in sworn statements. Given that there was no physical evidence identifying Davis as the gunman, the state’s case at trial hinged on these witnesses. Relied upon by the prosecution as credible witnesses in 1991, however, the state today portrays their revised testimony or recantations as “untrustworthy” and “unreliable”.

Judge Moore characterized the Troy Davis case as one in which “the evidence heard at trial was incomplete in some key manner”, adding that the new evidence did “not nullify the existence of the prior evidence”. He noted that “courts look upon recantation evidence with suspicion” and a court’s “general antipathy towards affidavit testimony counts double where the affiant is submitted in lieu of live testimony to prevent cross-examination and credibility determinations”.

Some, but not all, of the witnesses who have signed post-conviction affidavits appeared at the evidentiary hearing Judge Moore held in his court on 23 and 24 June 2010. One, Harriet Murray, has died since signing her statement in 2002. Others were available but were not called by the defence. Judge Moore reserved particular criticism for the failure of the defence lawyers to subpoena Sylvester Coles, the alternative suspect in the case to whom some of the affidavits pointed. The judge suggested that the lawyers “appeared to forget that the witness stand is the crucible of credibility”, and their “reluctance to put Mr Coles to the test robbed the Court of its ability to accurately assess Mr Coles’s claim that he did not shoot Office MacPhail.” The defence has described in an affidavit how they attempted to serve a subpoena on Sylvester Coles on the morning of 24 June 2010, both at his home and his place of employment. Judge Moore dismissed their efforts as “half-hearted” and “eleventh hour”.

Judge Moore ruled that some of the witness evidence pointing to Sylvester Coles as the gunman or alleging that the witness in question had felt pressured or coerced to testify against Davis was not credible; other evidence was “too general to provide anything but smoke and mirrors”; and yet other witness evidence carried little weight, either because it amounted to hearsay or because the witness was not presented at the evidentiary hearing. One such person was Dorothy Ferrell. At the 1991 trial, she had identified Troy Davis as the gunman. In an affidavit signed nine years later, she retracted this testimony and said that she had not seen who shot Officer MacPhail and that her testimony had been coerced. Dorothy Ferrell’s affidavit was “a clear recantation”, Judge Moore said. However, the fact that Troy Davis’s lawyers had not called her to the witness stand at the evidentiary hearing “destroys nearly its entire value”. Judge Moore described the decision not to call her to the stand as “especially curious because, based upon the contents of her affidavit and her lack of any obvious connections to Mr Davis, it would appear she should have been his star witness”.

Judge Moore decided that only one of the witness recantations amounted to a “meaningful, credible recantation”. This was Kevin McQueen’s. In September and October 1989, he had been held in the same jail as Troy Davis and he told police that Davis had confessed to him that he had shot Officer MacPhail. In 1996, McQueen signed an affidavit that he had made this up, and at the June 2010 hearing in Judge Moore’s court he reiterated that there had been “no truth” to his trial testimony against Davis.

Judge Moore concluded that Kevin McQueen’s trial testimony was indeed false. However, he found that his retraction “only minimally reduces” the state’s case against Troy Davis, suggesting that the fact that “Mr McQueen’s trial testimony was so clearly fabricated” should have been apparent to everyone at the time of trial, including the jurors. As such, he said “it is hard to believe Mr McQueen’s testimony at trial was important to the conviction”. The state appeared to consider it important, however – as Judge Moore himself pointed out, “the State persists in trying to support its veracity”.

Given that Judge Moore tasked himself with making “a probabilistic determination about what reasonable, properly instructed jurors would do”, and decided that Kevin McQueen’s false testimony would have had little impact on the jury’s verdict, it is worth reflecting on the compelling evidence that US capital jurors are more ready to believe what the prosecution presents to them than would those individuals who are systematically excluded under US law from serving on capital juries.

A US capital juror is not representative of the wider community in a country in which, as recently retired US Supreme Court Justice John Paul Stevens pointed out in 2007, “millions of Americans oppose the death penalty”. At jury selection, the defence and prosecution will question prospective jurors and have the right to exclude individuals, either for a stated reason (for cause) or without giving a reason (a peremptory challenge). Those citizens who would be “irrevocably committed” to vote against the death penalty can be excluded for cause by the prosecution, under a 1968 US Supreme Court ruling. In 1985, the Supreme Court expanded the class of potential jurors who could be dismissed for cause during jury selection. Under this standard, a prospective juror can be dismissed for cause if his or her feelings about the death penalty would “prevent or substantially impair the performance of his duties as a juror in accordance with his instructions and his oath”.

As early as 1986, the US Supreme Court acknowledged evidence from numerous studies that the “death qualification” of capital jurors “produces juries somewhat more ‘conviction-prone’ than ‘non-death-qualified’ juries”. Three Justices referred to the “overwhelming evidence that death-qualified juries are substantially more likely to convict or to convict on more serious charges than juries on which unalterable opponents of capital punishment are permitted to serve”. They added that “death-qualified jurors are, for example, more likely to believe that a defendant’s failure to testify is indicative of his guilt, more hostile to the insanity defence, more mistrustful of defence attorneys, and less concerned about the danger of erroneous convictions” (emphasis added).

In 1998, a review of existing research indicated that a “favourable attitude towards the death penalty translates into a 44 per cent increase in the probability of a juror favouring conviction”. Another expert review in 1998 concluded that: “death-qualified jurors, regardless of the standard, are more conviction-prone, less concerned with due process, and they are more inclined to believe the prosecution than are excludable jurors.” Justice Stevens revisited this question in 2008, asserting that “the process of obtaining a ‘death qualified jury’ is really a procedure that has the purpose and effect of obtaining a jury that is biased in favour of conviction.”

When considering what a “reasonable juror” would do in this case, the question also arises as to whether the concept of a “reasonable juror” in 2010 is the same as that in 1991, the year that Troy Davis was tried. And a question that Judge Moore did not (and was not asked to) address is, even if a “death qualified” jury were to vote to convict Troy Davis again today, would it vote for a death sentence?

In line with a softening in public and political support for the death penalty in the USA over this period, capital jurors are today displaying a much greater reluctance to pass death sentences than they were two decades ago. In 1991, death sentencing rates in the USA were approaching their zenith. Some 268 people were sentenced to death in the country that year. Death sentencing would peak in the next few years – reaching an apex of 317 new death sentences in 1996 – before beginning to drop off. In 2008 and 2009, for example, there were 111 and 106 new death sentences respectively – each far fewer than half of the 1991 total. This decline has been reflected in Georgia also. From 2000 to 2008, Georgia passed an average of 2.3 death sentences per year, while from 1991 to 1999, the average was 8.2, and from 1982 to 1990, the average was just over 10. Across the country, it seems that a greater awareness of the possibility of irrevocable mistakes, coupled with increased confidence that public safety can be ensured by life imprisonment without parole rather than execution, has contributed to this change.

At least some of the jurors from Troy Davis’s 1991 trial have indicated that they would no longer vote for a death sentence. For example, one of the jurors said in 2007 that the post-conviction evidence left him with “some lingering doubt about Mr Davis’ guilt. In light of this doubt, I would not have sentenced him to death… I recommend that his death sentence be commuted to life.”This takes us back to the question of clemency, a power invested in the executive that the Chief Justice of the US Supreme Court described in 1993 as the “fail safe” against mistakes made in the “fallible” criminal justice system. In his ruling in the Troy Davis case 17 years later, Judge Moore noted that while it might have once been believed that “any serious showing of innocence would result in state relief by clemency or state judicial process”, that is, “the state would always admit its mistake and rectify it”, events since 1993 “shatter the notion of a perfect ‘fail safe’ system for truly persuasive proof of innocence”.

Yet at the same time, Judge Moore laid responsibility for determining whether society can and should live with the combination of an absence of judicial certainty on the one hand and an irrevocable punishment on the other, at the feet of the political branches of government and the electorate, not the judiciary:

“If state prosecutors in Georgia are comfortable seeking the death penalty in cases of heinous crimes where their proof creates less than an absolute certainty of guilt, and the people of Georgia, through their validly enacted laws allow such a system knowing that it may occasionally result in the erroneous imposition of punishment, [US Supreme Court precedent] suggests that the Constitution will not interfere.”

The Constitution, Judge Moore had concluded before turning to the question of whether Troy Davis had met the “extraordinarily high” burden to prove his innocence, prohibits the execution of an innocent person. Applying the US Supreme Court’s “evolving standard of decency” standard, the judge concluded that “objective indicia of societal standards indicate a consensus that the execution of innocent convicts should be prohibited, whether that innocence is proved before or after trial.”

In US Supreme Court decisions based on its “evolving standards of decency” analysis, such as the 2002 and 2005 rulings prohibiting the execution of offenders with “mental retardation” or who were under 18 years old at the time of the crime, the Court referred to international practice in support of the prohibition. Judge Moore made no reference to the evolving trend towards global abolition of the death penalty, with 139 countries abolitionist in law or practice today, a clear majority. These countries reject the judicial killing of anyone, let alone someone whose guilt is in doubt. One of the reasons countries abandon the death penalty is their recognition that “despite the best efforts of society”, the criminal justice system makes mistakes.

While international law is abolitionist in outlook, it recognizes that some countries still retain the death penalty. Pending abolition, the international community has agreed that certain safeguards must be met in capital cases. One of these safeguards concerns the burden of proof on the death penalty state:

“Capital punishment may be imposed only when the guilt of the person charged is based upon clear and convincing evidence leaving no room for an alternative explanation of the facts”.

Clearly, given the mistakes that have been shown to occur before and at trial, this standard should apply beyond that stage in a case. Indeed, in reporting to the United Nations under this safeguard, governments have pointed to the post-conviction reversal of death sentences in cases where there was doubt over the condemned prisoner’s guilt after the trial.

The burden of proof chosen by Judge Moore is much less protective of a condemned prisoner whose guilt is in doubt than human rights safeguards require. If the state’s case is not “ironclad”, to borrow Judge Moore’s description, it means there is room for an alternative explanation of the facts. The Board of Pardons and Paroles should commute Troy Davis’s death sentence. After all, it said in this case in 2007 that it would not allow an execution to go ahead “unless and until its members are convinced that there is no doubt as to the guilt of the accused”. The Board does not have to apply the “extraordinarily high” burden adopted by Judge Moore. And the power of executive clemency must assume its role as a genuine “failsafe”.

Doubt still exists. This should be enough for even a death penalty supporter to oppose the irrevocable step of execution.

For further information, see:

USA: ‘Unconscionable and unconstitutional’: Troy Davis facing fourth execution date in two years, May 2009, http://www.amnesty.org/en/library/info/AMR51/069/2009/en

USA: Room for doubt, no room for execution, 13 July 2010, http://www.amnesty.org/en/library/info/AMR51/060/2010/en


AI Index: AMR 51/077/2010 Amnesty International 27 August 2010