Imagine being locked up on death row for 20 long years with no physical evidence linking you to the crime. You’ve been convicted thanks to “eyewitnesses,” most of whom have changed their stories. How would you feel as you face the executioner?

Troy Davis knows. He was sentenced to death in 1991 for the killing a Savannah, Georgia, police officer. No murder weapon was found. Most witnesses who implicated him have now recanted, save one who many believe is the actual killer. Davis has continued to maintain his innocence while facing, and surviving, three execution dates.

In June, the U.S. Supreme Court mandated a hearing featuring some of these witnesses, but the judge presiding over the hearing ruled that Troy Davis had not proven his innocence. The judge wrote in his decision it would be wrong to execute an innocent person–something that, unbelievably, the U.S. Supreme Court has never definitely stated–but that in this type of hearing innocence had to be proven to an “extraordinarily high” degree, with “clear and convincing evidence.” Troy Davis could not clear that hurdle.

I attended Troy Davis’ hearing on June 23. The utter confusion of conflicting stories, testimonies, and statements made it virtually impossible to find clarity, let alone innocence. I watched as four witnesses admitted they lied at trial, four witnesses implicated another man as the killer, and three witnesses described police coercion during questioning. The next day, many of the state’s witnesses told a contrary tale. Without any solid physical evidence available, which could have been scientifically tested, it was impossible for anything to seem “clear and convincing.” So it’s not surprising that Davis couldn’t prove his innocence to the level the judge required–but it would be an outrage if Georgia were to execute him.

Eyewitnesses are notoriously unreliable. According to the Innocence Project, more than 75 percent of wrongful convictions are due, at least in part, to faulty eyewitness testimony. Witness testimony is open to interpretation and “he said/she said” guesswork. It leaves a lot of room for doubt. So what happens when the only evidence available is witness testimony? Can you pass the judge’s “clear and convincing” test under those circumstances? Perhaps, if your witnesses happen to be extremely reliable. But for Davis, the very same witnesses whose credibility led to convicting him in 1991 were found to be unreliable in 2010. The bottom line: there is no guarantee that the wrongly convicted will have “clear and convincing” evidence of their innocence.

Following Troy Davis’ first clemency hearing in July 2007, the Georgia Board of Pardons and Paroles declared that they would “not allow an execution to proceed in this State unless and until its members are convinced that there is no doubt as to the guilt of the accused.” In September of this year, Governor Ted Strickland of Ohio used the same reasoning in commuting Kevin Keith’s death sentence, stating while he believed “it is far more likely that Mr. Keith committed these murders,” he would grant clemency due to “real and unanswered questions” in his case.

This is the commonsense approach if we want to prevent wrongful executions. Georgia should take heed as a possible fourth execution date looms on the horizon for Troy Davis. If we really want to avoid putting the innocent to death, we should insist on removing all doubt about guilt, rather than demanding “clear and convincing” proof of innocence.

Print Friendly, PDF & Email
Larry Cox

Larry Cox is the executive director of Amnesty International USA.

OtherWords commentaries are free to re-publish in print and online — all it takes is a simple attribution to To get a roundup of our work each Wednesday, sign up for our free weekly newsletter here.

(Note: Images credited to Getty or Shutterstock are not covered by our Creative Commons license. Please license these separately if you wish to use them.)