Skip to main content
Find a Lawyer
Michael C. Dorf

Did the Supreme Court Recognize an Innocent Person's Right Not to Be Executed?


Wednesday, August 26, 2009

Twenty years ago, off-duty police officer Mark McPhail was shot and killed in a Savannah, Georgia parking lot. Based on information provided by Sylvester Coles, the police sought Troy Davis for the murder. Davis turned himself in and was charged with the crime. He was found guilty and sentenced to death based on the testimony of eyewitnesses.

Since then, however, nearly all of those witnesses have recanted, claiming in affidavits that they were pressured by police to name Davis as the perpetrator. Meanwhile, additional evidence has been found indicating that Coles, the prosecution's star witness against Davis, was the actual killer. Yet despite national and international attention--including pleas by former Georgia Governor and U.S. President Jimmy Carter, former Georgia Republican Congressman and federal prosecutor Bob Barr, and even Pope Benedict--neither the Georgia courts nor the Georgia Pardons and Parole Board has seen fit to stop Davis's execution.

Last week, the Supreme Court offered Davis a ray of hope. In response to his petition for a writ of habeas corpus, the Justices ordered that a federal district court in Georgia "should receive testimony and make findings of fact as to whether evidence that could not have been obtained at the time of trial clearly establishes [Davis's] innocence."

The Court's order in Davis was not unanimous, however. Justice Scalia, joined by Justice Thomas, dissented. Justice Scalia said that even if the district court were to find Davis to be innocent, there would still be nothing unlawful about executing him.

That startling position prompted a rebuttal by Justice Stevens, joined by Justices Ginsburg and Breyer. Justice Sotomayor did not participate in the case. We know that the order required a majority of participating Justices, or in this case, five out of eight. Thus, at least two of the remaining three Justices--Chief Justice Roberts and Justices Kennedy and Alito--must have agreed with Stevens, Ginsburg and Breyer that an evidentiary hearing should be held.

We do not know, however, whether all three agreed or if not, which two agreed, nor do we know whether, or to what extent, these other Justices in the majority shared the views expressed by Justice Stevens.

Although the Supreme Court's order and the separate opinions in Davis are terse, they raise two fundamental constitutional questions: First, does the Constitution forbid the execution of an innocent person who was convicted and sentenced to death after a trial that was free of constitutional error but that nonetheless led to an erroneous verdict? Second, if the answer to that first question is yes, then is there also a constitutional right to make the showing of innocence before a federal court in a habeas corpus action?

In sending the Davis case to the district court for an evidentiary hearing, the Supreme Court could be tacitly signaling that it thinks the answer to both of these questions is yes. However, as I shall explain in the balance of this column, that interpretation of the Court's action is not by any means certain to be correct. Shocking as it sounds, Davis could end up proving his innocence and still be put to death.

Is There a Right of the Innocent Not to be Executed?

Many of the protections established by the Bill of Rights, the Fourteenth Amendment, and statutory and judge-made rules of criminal procedure are specifically designed to prevent the conviction (and by extension, the execution) of an innocent person. The requirements that defendants be able to confront and cross-examine their accusers, that the government prove its case beyond a reasonable doubt, and that a jury unanimously find guilt, along with further protections for defendants, all combine to reduce the odds that justice so badly miscarries as to result in the conviction of an innocent. Additional procedural safeguards limit the application of the death penalty.

Nonetheless, the system is imperfect. Even apart from police or prosecutorial misconduct, innocent people are sometimes convicted. As the work of organizations like the Innocence Project reveals, DNA testing and other methods can exonerate those who have been wrongly convicted of capital (and other) offenses. And while there is debate over whether our system of appeals and post-conviction review has led to the actual execution of innocent people (and if so, how many), there is no doubt that the system could permit an innocent person to be executed.

In 1993, in Herrera v. Collins, the Supreme Court raised, but did not ultimately decide, the question whether it would violate the Constitution to execute an actually innocent person. Acknowledging an "elemental appeal" to the claim that the Constitution forbids executing the innocent, the Court nonetheless left open the question whether, "in a capital case a truly persuasive demonstration of ‘actual innocence' made after trial would render the execution of a defendant unconstitutional." Even if such a demonstration would indeed render an execution constitutionally forbidden, moreover, the late Chief Justice Rehnquist said for the Court, "the threshold showing for such an assumed right would necessarily be extraordinarily high." Finding that Herrera's proffered evidence did not satisfy this standard, the Court denied relief in that case.

Does the Supreme Court's order in Davis indicate that it has now resolved the question that it left open in Herrera? Possibly--but very possibly not. Chief Justice Roberts has often invoked the principle that the Court should, whenever possible, avoid deciding hard constitutional questions. The Davis remand could be seen as an application of that principle: If Davis can show his innocence before the district court, then the Supreme Court can decide whether it is nonetheless permissible for the state to execute him. But if there is no showing of innocence, that latter question will be left for another day.

Still, the Davis order is at least suggestive of the existence of a substantive constitutional right. It will be all the more so in the event that the district court finds Davis innocent and the Supreme Court leaves that result in place. If so, Davis will likely be understood by the lower courts to establish a constitutional right not to be executed for anyone who, notwithstanding his conviction of a capital offense, can present evidence that "clearly establishes" his innocence.

Though the phrase "clearly establishes" is left undefined in the Court's order, it would not be surprising if this language came to mean that innocence must be shown by "clear and convincing" evidence. "Clear and convincing evidence" is a fairly demanding burden of proof, used in some special circumstances, that falls somewhere in between the "more probable than not" test of most civil cases and the "beyond a reasonable doubt" test of criminal cases.

The "New Rule" Twist: Justice Scalia's Argument and Justice Stevens's Response

In his dissent in Davis, Justice Scalia argues that even if the Court were now to hold that there is a right of the innocent not to be executed, Davis cannot take advantage of that rule. Why not? Because a provision of the Anti-Terrorism and Effective Death Penalty Act ("AEDPA") authorizes federal courts to grant habeas relief only when continuation of the challenged detention or execution of the challenged sentence would violate already "clearly established Federal law, as determined by the Supreme Court of the United States." The Court in Herrera expressly left open in prior rulings the question whether there is a right of the innocent not to be executed, Justice Scalia says--and if the Court intentionally failed to speak on this point, then surely the relevant law cannot possibly be said to be clearly established.

Justice Stevens responds to Scalia's argument, however, by noting that Davis had filed his habeas petition within the original jurisdiction of the Supreme Court, rather than as an appeal of a lower court ruling. And, he says, a 1996 case, Felker v. Turpin, itself left open another question: the question whether AEDPA limits the Supreme Court's original jurisdiction.

Three Problems with Justice Stevens's Response

But Justice Stevens's response is triply peculiar. First, the relevant text of AEDPA does not distinguish between cases that are filed initially in the lower federal courts and those that are originally filed in the U.S. Supreme Court.

Second, the Felker Court did not appear to leave the "new rule" issue open. In Felker, the Court held that one of AEDPA's restrictions on appeals did not apply in cases filed originally in the Supreme Court, because the relevant statutory restriction was by its terms applicable only to petitions "filed in the district court." The majority then distinguished AEDPA's rules governing repetitive and new petitions, remarking that "these restrictions apply without qualification." In saying that the Felker Court left the "new rule" issue unresolved, Justice Stevens must be relying on the Felker majority's next statement: "Whether or not we are bound by these restrictions, they certainly inform our consideration of original habeas petitions." In light of what immediately preceded it, however, the "whether or not" language is difficult to understand as expressing true agnosticism. Indeed, the Felker Court also said that its "authority to grant habeas relief to state prisoners is limited by §2254, which specifies the conditions under which such relief may be granted," including the limitation on "new rules."

Third, and apart from these difficulties in how Justice Stevens would read AEDPA and the Felker case, characterizing the Davis case as involving an original habeas petition in the Supreme Court is only formally accurate, because the real action will only begin to get under way in the district court. More broadly, AEDPA's limits would be largely pointless if they could be circumvented by a habeas petitioner's simply filing first in the Supreme Court, and then having the case transferred to a district court in which AEDPA would have otherwise barred relief in an original action.

Nor is it sensible for the Justices to make a discretionary determination of which habeas cases the district courts should hear without the restrictions of AEDPA. Given the pyramidal structure of the federal court system, and the thousands of state prisoners who can and do file federal habeas petitions, such a top-down screening practice would be backwards.

A Better Argument By Justice Stevens Regarding AEDPA and Actual Innocence

Justice Stevens makes a much better point when he suggests that the AEDPA prohibition on new rules should be read to exempt actual innocence claims. AEDPA did not come out of nowhere. In part, it codified prior Supreme Court precedents, including the 1989 case of Teague v. Lane, in which a plurality adopted the prohibition on habeas courts announcing or applying "new rules." The core idea was that state courts that follow the law as it exists when they decide cases should not be punished (by having convicted prisoners released through habeas) for failing to anticipate subsequent changes in the law that were made by the federal courts after the state courts had already ruled.

However, the Teague Court said that the new-rule prohibition would not apply in two exceptional circumstances: First, habeas courts could announce and apply new rules placing certain kinds of primary conduct (such as adult consensual sodomy) wholly outside the power of the state to criminally proscribe. This exception was later broadened to include new rules making whole categories of people, such as the mentally retarded, ineligible for the death penalty.

Second, habeas courts could announce new "watershed rules of criminal procedure" (such as the right of indigent defendants to free counsel).

Justice Stevens obliquely suggests in Davis that the rule forbidding execution of the innocent falls into one or both of the Teague exceptions. It makes a whole category of people--the innocent--ineligible for the death penalty; and it is arguably a watershed rule of fundamental fairness.

The difficulty with this argument, however, is that when Congress wrote AEDPA, it codified the Teague prohibition on new rules, but not the Teague exceptions.

Some lower courts have said that this failure to codify was merely an oversight, and that Congress surely meant to include the Teague exceptions. But for self-described textualists like Justices Scalia and Thomas, that answer won't do at all. Indeed, they dismiss Justice Stevens's allusion to the tacit codification possibility in a footnote, saying that it does "not warrant a response."

A Constitutional Right to Habeas for the Innocent?

Yet Justice Stevens has one more arrow in his quiver. He also proposes that if the AEDPA limits apply to bar proof that a person sentenced to death is actually innocent, then those limits are themselves unconstitutional--at least in cases in which the state courts are unable or unwilling to hear new evidence of innocence. Although Justice Stevens does not explain what constitutional provision would be violated, there are two likely candidates.

One possibility is the Due Process Clause of the Fourteenth Amendment. Justice Stevens (and the Court) might be suggesting that it violates due process to deny an evidentiary hearing to a condemned prisoner who, if given such a hearing, could demonstrate his innocence.

The second possibility is that Justice Stevens thinks that applying AEDPA to bar a hearing to Davis would violate the Suspension Clause. Last year, in Boumediene v. Bush, the Court ruled that the Suspension Clause--which limits the circumstances under which Congress may suspend the privilege of the writ of habeas corpus--confers an affirmative right of Guantanamo Bay prisoners to habeas corpus (or an equivalent substitute).

Boumediene was a hard case because it involved habeas corpus for an alien held outside the formal territory of the United States. But in a different sense, it was an easy case: Habeas historically was a mechanism for challenging executive detention.

By contrast, although Davis is a citizen subject to sentence within the United States, he would be arguing that there is a constitutional right to habeas as a post-conviction remedy. Yet, as Justice Scalia notes in his Davis dissent, prior to the Civil War, federal habeas was not available to challenge state court criminal sentences. And it was not until the Twentieth Century that federal courts began to use the writ to enforce substantive constitutional rights, as opposed to merely testing the jurisdiction of the state courts.

Thus, in Davis the real dividing line between, on one hand, Justices Scalia and Thomas, and on the other hand, Justices Stevens, Ginsburg, and Breyer, may be a familiar one: whether the meaning of a constitutional provision--here, the right to habeas--is fixed by the practice at the time of adoption or evolves over time as practices and values change.

That said, it must be remembered that nearly all of the foregoing analysis is highly speculative. Only five Justices wrote or joined any opinion in Davis explaining their reasoning, and even the two written opinions are more suggestive than definitive.

Thus, the Davis case may portend recognition of the right of death-sentenced prisoners to present new evidence of innocence, but it may not. Whether the hearing the Court has ordered proves to be a "fool's errand," as Justice Scalia characterizes it, could depend on what the full Court says about the matter if and when the case returns to its docket.

Michael C. Dorf, a FindLaw columnist is the Robert S. Stevens Professor of Law at Cornell University. He is the author of No Litmus Test: Law Versus Politics in the Twenty-First Century and he blogs at

Was this helpful?

Copied to clipboard