Resistance and Recantation: Ruminations on the Execution of Troy Davis

Professor John Blume is a distinguished member of our renowned faculty at Cornell Law School.Professor Blume joined the faculty in 1993, while serving still serving as the Executive Director of the South Carolina Death Penalty Project. He has served since 1996 on the Habeas Assistance and Training Project Counsel, which consults the Defender Services Committee of the United States Courts.After joining the law school faculty, Professor Blume, in conjunction with Professors Eisenberg, Johnson, and Garvey, formed the “Cornell Death Penalty Project.” The project is dedicated to the neutral examination of how the death penalty operates.The Project runs three law school clinics, the Capital Trial Clinic, the Capital Post-Conviction Clinic, and the Capital Appellate Clinic. With the assistance of clinic students, the Project has represented approximately thirty death-row inmates and seven persons charged with capital crimes. In this post, Professor Blume argues that that the resistance of those involved in the initial prosecution of defendants to admitting that mistakes can be made and the suspicion of recanted testimony ultimately led to the execution of Troy Davis, even though substantial doubt was raised as to his guilt.

“Well first of all, I’d like to address the McPhail family. I’d like to let you all know, despite the situation – I know all of you are still convinced that I am the person that killed your father, your son and your brother, but I am innocent. The incident that happened that night was not my fault. I did not have a gun that night. I did not shoot your family member. But I am sorry for your loss. I really am – sincerely. All I can ask is that each of you look deeper into this case so that you really will finally see the truth. I ask my family and friends that you all continue to pray, that you all continue to forgive. Continue to fight this fight. For those about to take my life, God have mercy on all of your souls. God bless you all.”

Troy Davis, September 21, 2011

“I can’t believe that it’s really happened. All of the feelings of relief and peace I’ve been waiting for all these years, they will come later. I certainly do want some peace. You know [Troy Davis has] been telling himself that [he’s innocent] for 22 years. You know how it is, he can talk himself into anything. . . . I will grieve for the Davis family because now they’re going to understand our pain and our hurt. My prayers go out to them. I have been praying for them all these years. And I pray there will be some peace along the way for them.”

Joan McPhail Harris, September 21, 2011


Troy Davis was executed by the State of Georgia on September 21, 2011 for the 1989 murder of Mark McPhail, an off-duty police officer who was shot and killed while trying to defend a homeless man who was being assaulted in a parking lot.  Davis went to his death maintaining, as he had at trial and throughout all of his appeals, including the unprecedented federal evidentiary hearing ordered by the United States Supreme Court,[1] that he was not the person who shot McPhail.  Davis, of course, is not the only person who has been executed despite his protestations of innocence.[2]  Some who have maintained innocence to the end turned out to be guilty,[3] while others were almost certainly innocent.[4]  But the truth is,in most cases, we will never know for certain.  Davis’ case falls in the latter category.  The prosecution’s case against Troy Davis consisted of eyewitness testimony and jailhouse informants.  While most of the witnesses later recanted their testimony, it was not enough to persuade the state or federal courts, or the Georgia Board of Pardons and Parole, to spare his life.

But Why?  Post-trial developments certainly raised substantial doubt about Davis’ guilt.  His supporters rallying cry of “too much doubt” resonated with thousands of people in the United States and around the world.  Why did no one intervene?  Davis’ appeals were unsuccessful for two primary reasons: resistance and recantation.

By resistance, I am referring to resistance by police officers, prosecutors and judges to claims of actual innocence.[5]  No matter how clear the proof of innocence, those involved in the initial prosecution are, by and large, unwilling to admit a mistake was made.  There have been a number of cases where post-trial DNA testing conclusively demonstrated the defendant’s innocence, but the police officers and prosecutors who sent the person to prison or death row have insisted that the defendant was guilty.  For example, Earl Washington, a Virginia death row inmate, was exonerated in 2001.  Despite the fact that a very suspect confession from a mentally retarded defendant was the only evidence linking him to a rape-murder, Washington was found guilty and sentenced to death.  Subsequent DNA testing, agreed to by both the prosecution and the defense, established that Washington was not the contributor of the semen found on the victim’s body.  Instead of acknowledging Washington’s innocence, however, the prosecutors changed theories and argued that Washington must have had an accomplice (‘the unindicted ejaculator’) who committed the rape and assisted Washington in committing the murder.

Judges too have been unwilling to acknowledge that the system made a mistake and convicted an innocent person (especially of a capital crime).  Again, Earl Washington’s case is instructive.  The Fourth Circuit astonishingly rejected Washington’s constitutional challenge to his conviction concluding that it would not have made a difference to the jury had they known that the semen did not belong to Washington.[6]   The reasons for resistance are not entirely clear.  Some speculate, I believe correctly, that many actors in the criminal justice system simply cannot confront the fact of wrongful conviction and its implications for the ability of the system to obtain accurate results.[7]

By recantation, I am referring to the criminal justice system’s deep suspicion of witnesses who testify for the prosecution at trial and then subsequently claim that their trial testimony was not true.  The later claims of recanting witnesses, even under circumstances where there is reason to suspect the veracity of the trial testimony, are often dismissed.  In fact, there is a legal presumption that the original testimony is correct and that the recanted testimony is false.  This longstanding and deeply entrenched presumption, based on the idea that recanted evidence can upset society’s interest in the finality of judgments, is unreliable and may be given for suspect motives.[8]  Justice Breyer fell back on it when questioned about the Court’s failure to intervene in Davis’ case, stating “[t]he recanting of witnesses . . . is not enough.  It is necessary to have proof that someone else has had to pull the trigger.”[9]

Thus, despite the fact that the case against Davis rested almost entirely on two forms of notoriously unreliable evidence known to contribute to false convictions – eyewitness identification and jailhouse informants – resistance to the reality of wrongful convictions and suspicion of recanted testimony ultimately sealed Davis’ fate.

It can be plausibly argued that given the criminal justice system’s strong interest in finality, the presumption against crediting recanting witnesses is justified.  On the other hand, in Davis’ situation, where the evidence of guilt was based on such a shaky foundation, why should that be the case?  Why should the post-trial evidence provided by the trial witnesses be treated with more suspicion than the at least equally unreliable eyewitness evidence offered during Davis’ trial?[10]  Similarly, why should we be more dismissive of a jailhouse informant’s post-trial assertions that his trial testimony was false when it is widely known that most “snitches” are opportunistic liars?[11]  Here, the general presumption against recanted evidence should be relaxed because it stands at least on equal footing from a reliability perspective with the evidence of guilt.

While Davis’ post-trial evidence of innocence may not have been enough to stop his execution, it does appear to have further destabilized support for the death penalty in the United States.  A poll taken shortly after his execution showed that support for capital punishment dropped to 61%, the lowest level in nearly 40 years and a 3% drop from last year.[12]  Furthermore, only half of the American people believe the death penalty is applied fairly.[13]  The possible wrongful execution of Troy Davis, and the likely wrongful execution of Todd Willingham (as well as the 139 exonerations of persons convicted of murder and sentenced to death), have exposed fundamental flaws in a system which decides who will live and who will die.  Resistance to the idea of wrongful convictions and the suspicion of recanted testimony will not change that basic fact–And that may be the lasting legacy of Troy Davis.

[*]I would like to thank Beccy Helm for her research assistance.

[1]After exhausting the traditional state and federal post-conviction remedies, and after an execution date was set, Davis filed a petition for writ of habeas corpus in the original jurisdiction of the Supreme Court maintaining that his execution would violate the Eighth and Fourteenth Amendments because he was actually innocent.  Original habeas petitions, while not uncommon in capital litigation, are routinely denied summarily.  The Supreme Court, however, to the surprise of nearly everyone, transferred the case to the United States District Court for the Southern District of Georgia for a “hearing and determination . . . as to whether evidence that could not have been obtained at the time of trial clearly establishes petitioner’s innocence.”  In Re Troy Anthony Davis, 130 S.Ct. 1 (2009).   Following the hearing, the district court judge found that Davis had not met his burden of proof and the Supreme Court denied further review.  SeeIn Re Davis, No. CV409-130, 2010 WL 3385081 at *61 (S.D. Ga. Aug. 24, 2010).

[2]However, despite popular belief to the contrary, it is relatively rare; most death row inmates do not maintain their innocence.

[3]Roger Coleman was executed by the state of Virginia in 1992.  He adamantly denied that he was the perpetrator, and at the time of his execution, many believed that he was, in fact, innocent.  Posthumous DNA testing, however, implicated Coleman as the murderer. SeeMaria Glod and Michael D. Shear, DNA Tests Confirm Guilt of Executed Man, Wash. Post, Jan. 13, 2006,

[4]The case against Troy Willingham, who was executed by the state of Texas, has been thoroughly discredited.  See David Grann, Trial by Fire: Did Texas execute an innocent man, The New Yorker, Sept. 7, 2009,

[5]Daniel Medweed, The Zeal Deal: Prosecutorial Resistance to Post-Conviction Claims of Innocence, 84 B.U. L. Rev. 125, 181–182 (2004).

[6]Washington v. Murray, 4 F.3d 1285, 1290 (4th Cir. 1993).

[7]Aviva Orenstein, Facing the Unfaceable: Dealing with Prosecutorial Denial in Post-Conviction Cases of Actual Innocence, ­­__ San Diego L. Rev­­__. (forthcoming), available at

[8]See, e.g.,Haouri v. United States, 510 F.3d 350, 353 (2d Cir. 2007); see also People v. Caruso, 172 Misc. 191, 194 (NY. Sup. Ct. Kings County 1939) (“There is no form of proof so unreliable and untrustworthy as the testimony of a recanting witness.”).

[9]Laura Goldman, Justice Breyer Riffs on the Death Penalty, Citizens United, Bush v. Gore, Naked Philadelphian Blog (Oct. 21, 2011),

[10]Eyewitness identification evidence is the single greatest cause of wrongful convictions in the United States, and has been a contributing factor in more than 75% of the convictions of defendants subsequently exonerated by DNA testing.  See

[11]See Alexandra Natapoff, Beyond Unreliable: How Snitches Contribute to Wrongful Convictions, 37 Golden Gate U. L. Rev. 107 (2006-07) (“45.9 percent of documented wrongful capitalconvictions have been traced to false informant testimony, making snitches the leading cause of wrongful convictions in U.S capital cases”)

[12]Frank Newport, In U.S.,Support for the Death Penalty Falls to a 39-Year Low: Fifty-two percent say the death penalty is applied fairly, Gallup (Oct. 13, 2011),



  1. I can agree with Professor Blume that in some cases, people involved in the initial prosecution should be less resistant to admitting that mistakes can and have been made. However, I do believe that the courts should view recanted testimony with suspicion. Why should one trust the recantation of someone who originally stood up in court and testified under oath? The question is raised: when was he lying? I am inclined to believe that the original testimony should be given more weight. Years after a person testifies, he may have been subject to outside pressures that may have further distorted his memory of the events, or in the instance of death penalty cases, a person may feel pressured to recant knowing that someone’s pending death could be attributed to their testimony. The reasons for recanting testimony may actually be that the originally testimony was false. But, we can never be certain. Impermissible factors may influence a person’s decision to recant previous testimony and as a precaution to these types of factors, courts should view all recanted testimony with suspicion.

Comments are closed.