DPIC Special Report: The Innocence Epidemic

A Death Penalty Information Center Analysis of 185 Death-Row Exonerations Shows Most Wrongful Convictions Are Not Merely Accidental

The Biggest Dangers are Police and Prosecutorial Misconduct and Knowingly False Testimony

Posted on Feb 18, 2021

Introduction Up

In 1993, the House Judiciary Committee’s Subcommittee on Civil and Constitutional Rights conducted hearings on what was then a relatively unknown question: How significant was the risk that innocent people were being wrongfully convicted and sentenced to death in the United States. After taking testimony from four exonerees who had been wrongfully condemned to death row, Representative Don Edwards, the subcommittee chairman, asked the Death Penalty Information Center to research the issue and compile information on how frequently these miscarriages of justice were occurring and what were the reasons why.

DPIC began looking more closely into death-row exonerations in the U.S. in the twenty years since the Supreme Court ruled in Furman v. Georgia in 1972 that the death penalty as then administered was unconstitutionally arbitrary and capricious. That research — undertaken before the availability of the internet — uncovered 48 cases in which a wrongfully convicted person had been released from death row because of innocence. The results of DPIC’s research were released in a Staff Report by the Subcommittee on Civil and Constitutional Rights Committee on the Judiciary, One Hundred Third Congress, First Session, Innocence and the Death Penalty: Assessing the Danger of Mistaken Execution, issued on October 21, 1993, and became DPIC’s first Innocence List.

The original list of 48 cases included 43 in which the defendant had subsequently been acquitted or pardoned or all charges had been dropped. Three cases involved compromise resolutions in which innocent defendants were immediately released upon pleading guilty or no contest to a lesser offense. One of the other two defendants was released from prison after the parole board became convinced of his innocence, and the other was acquitted at a retrial of the capital charge but convicted of lesser related charges.

DPIC made the decision at that time to maintain as comprehensive an innocence list as possible going forward. However, because of the inherent subjectivity of declaring a person innocent when some facts may remain in dispute, DPIC has adopted the objective criterion of “legal exoneration” for an individual to be included. What that means is that individuals who had been wrongfully convicted and sentenced to death were:

  • Subsequently acquitted of all charges related to the crime that placed them on death row, either at retrial or by an appellate court determination that the evidence presented at trial was insufficient to convict;
  • Had all charges related to the crime that placed them on death row dismissed by the prosecution or had reprosecution barred by the court in circumstances implicating the reliability of the evidence of guilt; or
  • Been granted a complete pardon based on evidence of innocence.

This decision means that dozens of innocent individuals who had been coerced into entering pleas as a condition to obtaining their freedom after conviction for crimes they did not commit are not included in the list. It also means that individuals who are innocent of murder but still have a record of conviction for some offense related to the crime in which a person was killed are also not included on the list. As a result, the Innocence List is a conservative estimate that likely substantially understates the number of innocent people who have been wrongfully convicted and sentenced to death, and in some unknown and untrackable number of cases been wrongfully executed.

Since the House subcommittee’s release of DPIC’s initial Innocence List, the Death Penalty Information Center has been tracking new exonerations to add to the list and has occasionally removed cases from the list when new discoveries indicated that the individual had not been fully exonerated. During that time, the list has more than quadrupled from 43 to 174 death-row exonerations. However, DPIC had not had the opportunity to take a new look at old cases to see if there were exonerations we had not detected in our initial research.

In 2017, DPIC embarked on a multi-year project to ascertain the status of every death sentence imposed in the United States since the Furman decision. As part of this death-row census project, we have obtained and reviewed information from state departments of corrections, researchers, prosecuting and defense offices, court files and dockets, and news archives for more than 9,600 state, federal, and military death sentences imposed in the United States since July 1972. The project — the most ambitious to date in tracking modern U.S. death sentences — uncovered nearly two dozen older cases that at first glance appeared to qualify as death-row exonerations. Further research narrowed that number to the eleven cases we have now added to the Innocence List, expanding the list to 185 exonerations.

With these eleven prior exonerations added to the list, DPIC has undertaken an analysis of many of the geographic and demographic features of these cases and an examination of the factors contributing to wrongful capital convictions. The results are disturbing — not merely because of the frequency with which U.S. states wrongfully condemn the innocent to die but because of the reasons why.

It is widely acknowledged that any system that is run by human beings inevitably makes mistakes and that, despite our best efforts, innocent people will be sentenced to death. But our analysis of death-row exonerations shows that innocent people are sentenced to death in most states and in every region of the country that authorizes capital punishment. They can be sentenced to death anywhere, but are most often wrongfully condemned in states and counties with a history of aggressive pursuit of the death penalty or that authorize outlier practices that make it easier to impose capital sanctions.

Moreover, the data show that most wrongful capital convictions and death sentences are not merely accidental or the result of unintentional errors. Instead, they are overwhelmingly the product of police or prosecutorial misconduct or the presentation of knowingly false testimony. More likely than not, they involve a combination of the two.

The data show that for every 8.3 executions carried out in the United States, a wrongfully condemned death-row prisoner is exonerated. That is an appallingly and unacceptably high rate of error. And as the United States approaches the 50th anniversary of the Furman decision, it raises the fundamental question of whether we can trust our state and federal governments to fairly, honestly, and reliably carry out capital punishment.

Summary of Key Findings Up

Since states began reenacting capital punishment statutes in the wake of the Supreme Court’s 1972 decision in Furman v. Georgia striking down existing death penalty laws, at least 185 people who were wrongfully convicted and sentenced to death have been exonerated. These wrongful capital convictions have happened in 29 different states and in 118 different counties, showing that, in whatever part of the country they are tried, capital defendants face an inherent risk of wrongful conviction.

Florida has had the most death-row exonerations of any state, with 30 since 1973, followed by Illinois with 21, and Texas with 16. Cook County, Illinois leads all counties with the most death-row exonerations (15) since 1973, followed by Cuyahoga County, Ohio; and Philadelphia County, Pennsylvania, with six exonerations each. Maricopa County, Arizona; and Oklahoma County, Oklahoma had five each.

Those five counties, each with a history of police and prosecutorial misconduct and of being outliers in their excessive pursuit of the death penalty, account by themselves for a fifth (20%) of the nation’s death-row exonerations. And more than 95 percent of wrongful capital convictions and death sentences from those counties involved some combination of police or prosecutorial misconduct and witness perjury or false accusation.

26 counties have more than one death-row exoneration, collectively accounting for 50.3% of all of the wrongful capital convictions that have resulted in exonerations.

Other key findings of DPIC’s research include:

  • Of the 185 exonerations that have occurred since 1973, 69.2 percent (128) have included official misconduct by police, prosecutors, or other government officials. Official misconduct was much more likely in cases involving defendants of color, cases in which exonerations took two decades or more, and cases in which DNA evidence was a significant factor in proving innocence.
    • Misconduct was a factor in more than three-quarters of cases in which Black defendants were exonerated (78.8%), more than two-thirds of cases involving Latinx defendants (68.8%), and 58.2 percent of cases with white defendants.
    • Misconduct occurred in 55.6 percent (55 cases) of the 99 cases in which exoneration took a decade or less, rising to 81.1 percent (43 of 53 cases) in exonerations taking 11-20 years, 88.0 percent (22 of 25) in the cases in which exoneration took 21-30 years, and in all 8 of the exonerations that took more than three decades.
    • Misconduct was present in 85.7 percent of the cases in which DNA evidence contributed to proving a death-row exoneree’s innocence, suggesting that the denial of DNA testing or absence of DNA evidence has caused innocence to be undetected or contributed to the denial of relief in other innocence cases.
  • Two-thirds (125) of exoneration cases (67.6%) have included a false accusation or perjury. Like official misconduct, perjury or false accusation was more likely in cases involving defendants of color (70.7% of Black and 93.8% of Latinx exonerees), and in cases in which exonerations took longer (84.8% of cases in which exoneration took 21 or more years). False or misleading forensic evidence was present in 31.9 percent (59) of exoneration cases and false or fabricated confessions were implicated in 16.2 percent (30) of exonerations
  • Underscoring the often intentional nature of wrongful capital convictions, more than half of all exonerations (100 cases, 54.1%) involved both official misconduct and perjury or false accusation, and at least one or the other was present in 153 of the 185 exonerations (82.7%). Among counties with multiple wrongful capital convictions, at least one of these factors was present in 84 of the 93 death-row exonerations, or 90.3 percent of the time.
  • Exoneration took significantly longer for Black defendants who were wrongly convicted and sentenced to death. DPIC found that it took Black death-row exonerees an average of 4.3 years longer to be cleared than their white counterparts. African Americans have accounted for 12 of the 13 post-Furman wrongful convictions that have taken 30 years or longer to exonerate.
  • Outlier practices contribute to sending innocent people to death row. More than 15 percent of all death-row exonerations in the U.S. are in cases in which trial judges overruled jury recommendations for life or imposed the death penalty based on non-unanimous jury votes for death. At least 23 exonerations in Florida, five in Alabama, and one in Delaware involved this outlier practice.

Jurisdictions Up

In July 1972, the Supreme Court struck down all existing capital punishment statutes in the United States, ruling in Furman v. Georgia that the death penalty was being administered in an unconstitutionally arbitrary and capricious manner across the country. States immediately began reauthorizing capital punishment through statutes and state appeal procedures that were supposed to ensure that the death penalty would be carried out consistently and reliably. But even before the first execution took place under the new laws in January 1977, ten innocent men who had been wrongfully convicted and condemned under those laws had already been exonerated.

Since Furman, at least 185 people who were wrongfully convicted and sentenced to death have been exonerated in the United States. The wrongful convictions that have led to these exonerations have occurred in 118 different counties in 29 different states. The breadth and geographic arbitrariness of these exonerations demonstrate that wherever capital cases may be tried, innocent defendants face a risk of being wrongfully sent to death row.1

From Furman through February 15, 2021, 1,532 prisoners have been executed in the United States. Prior to DPIC’s death-row census project, it had been estimated that there had been one exoneration for every nine executions. The new data show that the ratio of exonerations to executions is even worse. For every 8.28 death-row prisoners who are executed, one former death-row prisoner who was wrongfully convicted and wrongfully sentenced to die has been exonerated.


Florida has the most documented death-row exonerations of any state, with 30 since 1973. Four of those exonerations are for convictions under pre-Furman death-penalty statutes. The state’s 26 exonerations in death sentences imposed since Furman are also the most in the country. Illinois is second in the number of exonerations with 21, followed by Texas (16), Louisiana and Ohio with 11 each, and Arizona, Oklahoma, and Pennsylvania with ten each.

Florida also has the most counties with death-row exonerations, with 20. Texas (13) and North Carolina (11) are the only other states to have wrongful capital convictions and death-row exonerations in more than ten counties. Georgia, Illinois, and Oklahoma have had wrongful capital convictions in six counties that have resulted in exonerations and Alabama, California, Louisiana, Mississippi, Ohio, and Pennsylvania have had death-row exonerations in five counties each.

Eighteen of Illinois’ 21 death-row exonerations (85.7%) have involved police or prosecutorial misconduct, the most misconduct-related exonerations in the nation. Its 19 death-row exonerations involving perjury or false accusation are also the worst in the nation. Florida has the next largest number of misconduct-related death row exonerations with 13, followed by Louisiana with 11. Ohio and Texas are next with ten each. Every death-row exoneration in Louisiana has involved official misconduct.

More than 90% of Illinois’ death-row exonerations involve wrongful convictions obtained as a result of perjury or false accusation. Two other states — Florida (17) and Texas (11) — have more than ten death-row exonerations involving perjury or false accusation, followed by North Carolina, Ohio, and Pennsylvania with nine each.

One significant contributing factor in Florida’s exonerations is the practice of allowing death sentences without a unanimous jury recommendation for death. Until 2016, Florida permitted trial judges to impose death sentences despite jury recommendations for life or based on non-unanimous jury votes for death. (See Outlier Practices, below.) Of the 25 Florida death-row exonerations for which the jury vote is known, 22 involved judges who imposed the death penalty by overriding a jury recommendation for life or following a non-unanimous jury recommendation for death.


Capital defendants in 118 U.S. counties have been exonerated after being wrongfully convicted and sentenced to death. While 92 counties have one exoneration each, the 26 counties with multiple exonerations collectively account for 50.3% of all of the wrongful capital convictions in the U.S. that have resulted in exonerations.

Fewer than a quarter of counties with wrongful capital convictions that have resulted in exonerations have had more than one exoneration (22.0%). However, the more exonerations a county has, the more likely the exonerations are to involve wrongful conduct rather than just human error.

The 26 counties with multiple death-row exonerations account for 63.3% of all of the exonerations involving false confessions, 57.0% of the exonerations with official misconduct, 56.8% of the exonerations involving false accusations or perjury, and 48.6% of the mistaken identification exonerations. 90.3% of the exonerations in these counties involve either official misconduct or perjury/false accusation or both.

Cook County (Chicago), Illinois wrongfully convicted and condemned 15 death-row exonerees since 1973, more than double the number of any other county in the United States. It is followed by Cuyahoga County (Cleveland), Ohio; and Philadelphia County, Pennsylvania, with six exonerations each. Maricopa County (Phoenix), Arizona; and Oklahoma County (Oklahoma City), Oklahoma had five each.

What these five counties have in common are a history of police and prosecutorial misconduct and their historically excessive pursuit of the death penalty. By themselves, they account for 37 death-row exonerations, one fifth (20%) of the nation’s total. And more than 95 percent of their wrongful capital convictions and death sentences involved some combination of police or prosecutorial misconduct and/or witness perjury or false accusation.

Because death-row exonerations represent one of the most glaring breakdowns of the legal system, a large number of exonerations marks a county as a true outlier. In their landmark 2000 study, A Broken System: Error Rates in Capital Cases, Columbia University law professor James Liebman and his colleagues found that the more aggressively a jurisdiction pursued the death penalty and obtained death sentences, the higher the percentage of cases in which courts found prejudicial constitutional error. That same phenomenon appears to be the case with exonerations from wrongful capital convictions.

DPIC’s review of the 185 exonerations found that high numbers of exonerations are correlated to heavy use of the death penalty. It is not coincidental that every county with five or more exonerations also appeared in DPIC’s October 2013 report, The 2% Death Penalty, among the 2% of U.S. counties responsible for more than half of the U.S. death-row population or for carrying out a majority of the nation’s executions. The link between disproportionally heavy use of the death penalty and official misconduct is well established, so it is not a surprise that the counties with the most exonerations also have disproportionally high rates of misconduct. Of the 37 exonerations originating in the top 5 counties, 32 (86.5%) involved official misconduct. Among all exonerations, 67.6% involved official misconduct.

Illinois’ Cook County (Chicago) stands out for its enormous number of exonerations. Its 15 death-row exonerations are directly related to endemic police corruption, as the notorious “Burge Squad,” operating under Chicago Police Commander Jon Burge, and disgraced Chicago detective Reynaldo Guevara systematically tortured or coerced innocent suspects into confessing to murders they did not commit. Illinois’ high rate of wrongful convictions in death cases was a major factor in the state’s 2011 repeal of capital punishment, as state officials decided there was no way to correct the inaccuracy of the state’s death penalty system.

The data on the causes of wrongful convictions in the Cook County death-row exonerations document the systemic misconduct that made the county a national example of corrupt police practices. All but one death-row exoneree in Cook County has been a person of color – 13 were Black and one was Latino. All of the exonerations involved either perjury or false accusation or official misconduct and the 12 with official misconduct involved both. 14 of the exonerations involved perjury or false accusations. Verneal Jimerson’s case is typical of the Chicago wrongful capital convictions. Jimerson was convicted on the basis of testimony that the witness later recanted, saying police had coerced her into lying. In exchange for her false testimony, prosecutors reduced her sentence from 50-years imprisonment to 2-years probation. Jimerson spent 11 years on death row before he was exonerated in 1996.

Nearly half (7) of Cook County’s exonerations also involved false confessions. One illustrative case is that of Gabriel Solache, who was sentenced to death in 2000 and exonerated in 2017. Solache and his co-defendant, Arturo DeLeon-Reyes, were beaten, deprived of sleep, and given little to eat or drink over three days of interrogation by Detective Guevara. Though he spoke only Spanish, the confession that Solache purportedly gave to police was written entirely in English. Guevara has been implicated in more than 50 wrongful murder convictions.

The two counties with the next most exonerations — Cuyahoga County (Cleveland), Ohio and Philadelphia County, Pennsylvania — have six exonerations each. In both counties, every exoneration involved official misconduct. As did Cook County, Philadelphia and Cuyahoga counties also have high rates of perjury or false accusation. In Philadelphia, half of death-row exonerations also involved inadequate legal defense, a by-product of the city’s well-documented failures in the selection and compensation of appointed counsel.

Wrongful capital convictions and prosecutions in Philadelphia go far beyond the six official death row exonerations. In the three years since Philadelphia District Attorney Larry Krasner established the office’s Conviction Integrity Unit, it has been instrumental in 18 exonerations, including those of death-row prisoners Christopher Williams and Walter Ogrod. Williams had been wrongfully charged with six different murders and was acquitted in two trials and exonerated of the other four murders.

A seventh likely death-row exoneree, Frederick Thomas, died before he could be released, while the Philadelphia District Attorney’s office, knowing Thomas had terminal cancer, fought defense efforts to expedite the prosecutors’ appeal of the court order that had granted him a new trial. Before Thomas was granted a new trial in 2002, the state’s two eyewitnesses recanted their testimony and police officer James Ryan — who the defense said had framed Thomas — was convicted on corruption charges arising out of his conduct in other cases, including falsifying police reports and making false arrests. Unknown even to his lawyers at the time of Thomas’ death — and withheld by Philadelphia prosecutors — Ryan and the same informant who testified against Thomas had also framed another innocent man in a different murder case.

James “Jimmy” Dennis, who also was wrongfully convicted of murder, spent more than 25 years on Pennsylvania’s death row. After his conviction was overturned for multiple instances of prosecutorial misconduct, he pled no contest to lesser charges to secure his release in 2017. Dennis, along with Anthony Wright and Percy St. George, were wrongfully capitally prosecuted in the early 1990s as a result of evidence fabricated or hidden by homicide detectives Manuel Santiago and Frank Jastrzembski.

Thomas and Dennis were both prosecuted by Assistant District Attorney Roger King, who at one point had been responsible for 20% of all the death sentences imposed in Pennsylvania. King also prosecuted death-row exoneree William Nieves, despite evidence that an eyewitness had described the shooter as being a different race and having a different body type than Nieves. King also attempted to prosecute four innocent men in Philadelphia’s “Lex Street Massacre,” the worst mass murder in the city’s history. No physical evidence linked any of the men to the killings, but King moved forward with one questionable witness and the coerced confession of one of the defendants. After 18 months in prison without being tried, the court dismissed all charges against the men.

The Cuyahoga County triple-exoneration of Ricky Jackson, Wiley Bridgeman, and Kwame Ajamu exemplifies the factors that contribute to wrongful convictions. The three men were convicted in 1975 on the testimony of a 12-year-old boy. Decades later, a journalist’s examination of the case prompted the witness to come forward and explain that he had lied to police, and when he attempted to recant his identification of the men before their trial, he was intimidated and coerced by police into maintaining his testimony. Jackson, Bridgeman, and Ajamu were exonerated in 2014, 39 years after their convictions.

Cuyahoga County prosecutors have a long history of misconduct in capital cases and of stonewalling efforts to release death-row prisoners who have been wrongfully convicted. Joe D’Ambrosio and Thomas Keenan were wrongfully convicted and sentenced to death in 1989 for the murder of Anthony Klann, whose body was found in a creek in a Cleveland park. No physical evidence linked either man to the crime. They were convicted based on the false testimony of Edward Espinoza, who pled guilty to lesser charges and served 12 years in prison in exchange for his testimony.

Both men overturned their convictions because of prosecutorial misconduct. A federal court granted D’Ambrosio a new trial in March 2006, after prison chaplain, Rev. Neil Kookoothe uncovered exculpatory evidence that that the Cuyahoga County Prosecutor’s office had concealed from the defense. That evidence included that Espinoza’s version of the murder was false, that Klann had not even been murdered at the creek, and that Klann may instead have been murdered by a man who had previously raped D’Ambrosio’s roommate and who then deflected attention from himself by falsely implicating D’Ambrosio in Klann’s murder. County prosecutors continued to withhold other exculpatory information from the defense, and then, a week before the trial, for the first time disclosed the existence of blood samples and soil samples. Then, after the trial was delayed, they waited for several more months before disclosing that Espinoza had died and so could not be cross examined about his false testimony.

In 2010, the federal court barred the county prosecutor’s office from retrying D’Ambrosio and he was finally exonerated in 2012 after prosecutors exhausted their appeals of that order. The prosecutor’s office then fought for eight more years to deny D’Ambrosio compensation for his wrongful conviction.

Keenan also overturned his 1989 conviction because of prosecutorial misconduct but was retried and resentenced to death in 1994. A federal court overturned that conviction as well, citing “egregious” prosecutorial misconduct. A state trial court then dismissed all charges against Keenan in 2012 and barred prosecutors from retrying him. However, county prosecutors successfully appealed that order and, to avoid facing a third capital trial, Keenan pled guilty to lesser charges in 2016. He was sentenced to time served.

Cuyahoga County prosecutors continue to stonewall other possible exonerations, fighting DNA testing and lying to the court about withholding evidence for decades in the case of Melvin Bonnell for the 1987 murder of Robert Bunner. In April 2020 — two months after Bonnell was scheduled to be executed — Bonnell’s lawyers found three envelopes containing bullets and shell casings in the prosecutor’s files. The Cuyahoga County Prosecutor’s Office had claimed for 30 years that the evidence had been lost or destroyed, continuing to tell the court in 2017 that they had no physical evidence from the case and that the boxes in which the defense ultimately found the bullets and shell casings “contained only paper documents.” Even after Bonnell’s lawyers had found the evidence, assistant prosecuting attorney Frank Zeleznikar lied to the court that “the items in question were not preserved for testing. … This has never been a secret. The State never hid it from Bonnell.”

Two more counties that are notorious for death-penalty abuses — Maricopa County (Phoenix), Arizona; and Oklahoma County (Oklahoma City), Oklahoma — have the next most death-row exonerations, with 5 each.

Maricopa County has an extensive history of misconduct in capital cases. Since 2012, two of the county’s prosecutors — Andrew Thomas and Juan Martinez — have been disbarred for misconduct. The county seeks the death penalty so often that in 2017 it ran out of defense attorneys qualified to handle newly-charged capital cases. One of the county’s death-row exonerees, Debra Milke, spent 23 years on death row before a court barred her retrial because of “egregious” police and prosecutorial misconduct.

Ray Krone, an honorably discharged Air Force veteran, was wrongfully convicted and sentenced to death in 1992 for the murder of a waitress in a Phoenix bar. The prosecution’s case against him rested on circumstantial evidence and the junk-science testimony of a later discredited expert witness who falsely testified that bite marks found on the victim matched Krone’s teeth, which had been disfigured in an accident. The Arizona Supreme Court overturned Krone’s conviction, finding that prosecutors had sandbagged the defense by withholding until the trial was underway a video exhibit purporting to demonstrate that Krone’s teeth matched the bitemark on the victim. He was retried and convicted again based upon the junk bitemark testimony.

Subsequently, the post-conviction court granted Krone’s request for DNA testing. The results exonerated Krone and when they were submitted to a national DNA databank, identified the actual killer.

In Oklahoma County, death-row prisoners Julius Jones and Richard Glossip face execution despite strong evidence of innocence. Both allege that they were wrongfully convicted under the administration of notorious prosecutor “Cowboy” Bob Macy. Macy sent 54 people to death row during a 21-year tenure as District Attorney that was marked by prosecutorial misconduct. Jones says that a combination of racial bias, poor representation, and false informant testimony led to him being sentenced to death for a crime he did not commit.

Glossip was sentenced to death for the 1996 murder of motel operator Barry Van Treese. No physical evidence linked Glossip to the murder, and the only evidence implicating him came from the multiple conflicting stories of the actual killer, Justin Sneed, a 19-year-old methamphetamine addict who was spared the death penalty in exchange for testifying that Glossip had offered to pay him to kill Van Treese.

Glossip’s lawyers charge that the Oklahoma County DA’s office has engaged in witness intimidation to impede Glossip’s attempts to prove his innocence. After two former prisoners who knew Sneed came forward with information that Sneed had acted alone and that it was common knowledge in the prison that Sneed had lied about Glossip to receive a reduced sentence, the defense says, current Oklahoma County District Attorney David Prater had arrest warrants issued for both men. Glossip came within hours of execution in September 2015, but the execution was halted when prison officials discovered that they had been provided the wrong drug to carry out the execution.

23 of Macy’s capital convictions relied heavily on the testimony of disgraced police chemist Joyce Gilchrist, who an FBI investigation in 2001 concluded had offered testimony “that went beyond the acceptable limits of science.” An internal police investigation found that evidence in many of Gilchrist’s major cases was missing, along with three years of her blood analysis files. In the case of Curtis McCarty, one of three death-row exonerees prosecuted under Macy, Gilchrist falsely testified that hairs found at the crime scene matched McCarty’s and that his blood type matched the semen found on the victim’s body. A later investigation revealed that Gilchrist had altered her notes to implicate McCarty and that the hairs she had tested were missing. McCarty was exonerated in 2007 after independent DNA testing excluded him as a suspect. Almost half of the 23 people who were sentenced to death in trials in which Gilchrist testified were executed before their cases could be reviewed.

Outlier Practices: Death Sentences Imposed Despite the Votes of One or More Jurors for Life Up

The Data — Exonerations in Florida, Alabama, and Delaware

Death sentences imposed by trial judges overriding jury votes for life or disregarding the life votes of individual jurors account for the vast majority of death-row exonerations in the three states — Florida, Alabama, and Delaware —that permitted these practices. Twenty-nine of the 32 exonerations in these states (90.6%) for death sentences imposed during a period in which the state permitted a death sentence based on a non-unanimous jury recommendation for death or the judicial override of a jury’s vote for life involved one or the other of these practices.

Florida. Florida’s 30 death-row exonerations since 1973 are the most in the nation, as are the 26 exonerations in cases tried since Furman. Allowing judges to impose death sentences despite one or more jurors voting for life is a major reason why. DPIC has been able to determine the jury votes in 25 of the 26 death-row exonerations involving Florida’s post-Furman judicial override/non-unanimity statute. In those 25 cases, juries unanimously recommended death only twice. 23 times, one or more jurors voted for life, including five instances in which judges overrode jury recommendations for life.

The practice also disproportionately affected defendants of color. Nearly seventy percent (16 of 23) of the wrongful capital convictions in which Florida judges imposed death sentences despite the votes of one or more jurors for life involved defendants of color. Eleven of these wrongful death sentences (47.8%) were imposed on Black death-row exonerees. Another five (21.7%) were imposed on Latinx exonerees. Seven white exonerees (30.4%) were also wrongfully sentenced to death by Florida judges under these practices.

Alabama. Seven former death-row prisoners have been exonerated in Alabama after post-Furman death sentences. At least one juror voted for life in five of the six exonerations in which sentencing juries were impaneled. In three of the cases, judges overrode jury recommendations for life—Larry Randal Padgett (9-3 jury vote for life); Daniel Wade Moore (8-4 for life); and Walter McMillian (7-5 for life). Two other wrongfully convicted death-row prisoners, Anthony Ray Hinton and Wesley Quick, were sentenced to death after non-unanimous sentencing recommendations by their juries. DPIC’s research suggests that the jury that wrongfully convicted Charles Bufford in 1978 unanimously recommended that he be sentenced to death. Alabama’s seventh death-row exoneree, Gary Drinkard, waived a jury sentencing altogether.

Delaware. There has been one post-Furman death-row exoneration in Delaware. Isaiah McCoy was sentenced to death by his trial judge following the jury’s 10-2 sentencing recommendation for death. Jermaine Wright, another likely innocent Delaware death-prisoner who was sentenced to death after a non-unanimous jury vote, won a new trial because of prosecutorial misconduct in his case. He was later released from custody upon pleading no contest to lesser charges.

Causes of Wrongful Convictions Up

DPIC’s analysis of the data from the 185 exonerations now on our Innocence List suggests that most innocent people who are wrongfully convicted and sentenced to death don’t end up on death row by mistake. Instead, the data show that far more frequently, and particularly with people of color, innocent death-row prisoners were convicted because of a combination of police or prosecutorial misconduct and perjury or other false testimony. An innocence case with a single contributing cause is the exception, not the rule. More than three-quarters of death-row exonerations have multiple types of errors.

Beyond mere human fallibility, the data raise serious questions as to whether the criminal legal system has the capacity or the will to institute reforms to prevent wrongful capital convictions. For while eyewitness and interrogation procedures, jury instructions and scientific evidence can be improved, no procedure or scientific advance can eliminate intentional misconduct. Of the 185 exonerations that have occurred since 1973, more than 80% have involved official misconduct by police, prosecutors, or other government officials or perjury/false accusation, and more than half have involved both.

Official misconduct and perjury/false accusation are by far the leading causes of wrongful capital convictions. 69.2% (128) of the death-row exonerations have included official misconduct by police, prosecutors, or other government officials. 67.6% (125) have included false accusation or perjury. One or the other was present in 153 of the 185 exonerations (82.7%) and they were both present in 100 of the cases (54.1%).

False or misleading forensic evidence is the next leading contributing factor in wrongful capital convictions, present in 31.9% (59) of wrongful convictions. Mistaken witness identifications — often accompanied by suggestive or manipulated identification procedures — were present in 20.0% (37) of the exonerations and false or fabricated confessions were present 16.2% of the time (30 cases). Finally, appellate courts found that exonerees had been convicted based on insufficient evidence in 9.2% (17) of the cases. In at least 47 cases — more than a quarter of the exonerations (25.4%), inadequate legal representation failed to present key defense evidence or failed to contest erroneous or improper prosecution evidence or argument.

Just as wrongful capital convictions were usually the product of intentional misconduct, they were rarely the product of a single isolated error. DPIC’s analysis of the causes of wrongful capital convictions found that the conviction rested on a single type of error in only 21.6% of the cases (40 cases), and 11 of those cases were exonerations based on the insufficiency of the evidence. 70 of the convictions (37.8%) were the product of two different types of wrongful factors, but most frequently, three or more factors contributed to the wrongful capital convictions (75 cases, 40.5%).

The data show — and individual cases illustrate — that the combination of these factors often go hand-in-hand as part of a prosecutorial strategy to convict. They include prosecutors offering incentives for cooperating witnesses to testify favorably to state, withholding information about the witness’s expectation of favorable treatment, and eliciting false testimony from the witness. Likewise, a codefendant who is threatened with potential capital charges has a strong incentive to implicate others in exchange for a favorable plea deal. In one of the recently discovered exonerations, Anthony Carey was sentenced to death in Mecklenburg County, North Carolina for the 1973 robbery and murder of a gas station employee. James Mitchell, the actual shooter, agreed to plead guilty and implicate four other people in the murder in exchange for a non-capital sentence. Carey’s conviction was overturned on appeal, and prosecutors dropped charges after Mitchell recanted his false testimony.

Government officials have also pressured witnesses to provide false testimony. In Alfred Brown’s case, his girlfriend truthfully testified before the grand jury that he had been on the phone with her in her apartment shortly before the murder, making it physically impossible for him to have gotten to the murder scene in time to have committed the crime. Harris County, Texas prosecutors obtained phone records in an attempt to refute her testimony. But when the records actually corroborated Brown’s alibi, assistant district attorney Dan Rizzo withheld those documents from the defense and jailed Brown’s girlfriend for seven weeks until she changed her testimony to falsely implicate him. In Cuyahoga County, Ohio, Kwame Ajamu, Wiley Bridgeman, and Ricky Jackson were convicted based on the testimony of a 12-year-old boy who was pressured by police to testify against them. Police threatened to jail the boy’s parents at a time when his mother was battling cancer. Police also fabricated evidence and withheld evidence of the men’s innocence.

Race and Innocence Up

Wrongful capital convictions are not race neutral. DPIC’s exoneration data shows that exonerees of color, and particularly those who are Black, are more likely to be victims of official misconduct and false accusation, more likely to be wrongfully convicted and condemned, and more likely to spend longer periods facing execution or under the continuing shadow of their wrongful conviction than white death-row exonerees.

This should come as no surprise. As we explained in our September 2020 report, Enduring Injustice: The Persistence of Racial Discrimination in the U.S. Death Penalty, implicit and explicit racial bias permeate every stage of a capital case, from policing practices, to arrest and clearance rates, to prosecutorial charging, to jury selection, to the representation provided by defense counsel, to sentencing and the determination of “deathworthiness,” and, as the new data show, to exonerating those who are wrongfully condemned to die.

But the systemic ordinariness of this discrimination should make the numbers more disturbing, not less. In the context of racial justice and the supposedly race-neutral administration of the law, it amounts to a statement that members of communities of color are more disposable and that, especially, innocent Black lives matter less.

As of October 1, 2020, an already disproportionate 57.9% of those on death row in the United States or facing capital retrials or resentencing proceedings were people of color. 41.6% were Black; 13.4% were Latinx. An even more racially disproportionate 63.8% of wrongfully convicted death-row exonerees are people of color, 53.5% of whom are Black.

The wrongful convictions of Black exonerees are also much less likely to have been accidental. DPIC’s exoneration data also shows that that official misconduct — defined as misconduct by police, prosecutors, or other government officials2 — is disproportionately prevalent in death-row exonerations of defendants of color, especially Black defendants. The same is true of perjury and false accusation, which very often occurs in tandem with government misconduct.

Official misconduct occurred in more than two-thirds of death-row exonerations (69.2%) and was present in a majority of cases irrespective of the exoneree’s race. However, misconduct occurred with significantly greater frequency if the exoneree was Black or Latinx. Official misconduct was a contributing factor in the wrongful convictions of 78.8% of Black death-row exonerees and 68.8% of Latinx death-row exonerees, as compared to 58.2% of exonerations of wrongfully convicted white death-row survivors.

Official misconduct occurred in 20.6% more cases with Black exonerees and 10.6% more cases with Latinx exonorees than in cases involving white exonerees. The odds that official misconduct contributed to a death-row exoneree’s wrongful capital conviction were 2.7 times greater if the exoneree was Black than if he or she was white, and 1.6 times greater if the exoneree was Latinx.3

Similarly, false accusation or perjury was a factor in more than two-thirds (125) of death-row exonerations (67.6%). Like official misconduct, perjury or false accusation was more prevalent in cases involving exonerees of color (72.9% of cases), affecting 70.7% of Black exonerees and 93.8% of Latinx exonerees. It also was a factor in the wrongful convictions of 58.2% of white exonerees, the same number as were affected by official misconduct.

Perjury/false accusation occurred at a rate that was 35.6 percentage points higher for Latinx exonerees and 12.5 percentage points higher for Black exonerees than for their white counterparts. The odds that perjury or false accusation contributed to a death-row exoneree’s wrongful capital conviction were 1.7 times greater if the exoneree was Black than if he or she was white, and 10.8 times greater if the exoneree was Latinx.4 However, the comparisons with respect to Latinx exonerees are less meaningful because of the relatively small sample size.

False or fabricated confessions and mistaken eyewitness identification, while less prevalent as factors contributing to wrongful capital convictions, also were disproportionately present in cases involving Black exonerees. A DPIC analysis found that 18.2% of Black death-row exonerees’ cases involved false or fabricated confessions, as compared with the cases of 18.8% of Latinx exonerees and 13.4% of white exonerees. Mistaken identification occurred in 26.3% of Black exonerees’ cases, in contrast to 13.4% of cases involving white exonerees (the same percentage as for false confessions) and in only 6.3% of cases with Latinx exonerees.

False or misleading forensic evidence was a factor in a greater number, but smaller percentage of wrongful capital convictions involving Black exonerees (29 cases, 29.3%) than in cases involving white exonerees (23 cases, 34.3%). It was present in seven cases (36.8%) involving other exonerees of color.

9.2% of exonerations involved appellate acquittals as a result of insufficient evidence. Appellate acquittals comprised 13.4% of exonerations involving white death-row exonerees and 12.5% of exonerations involving Latinx exonerees, but only 6.1% of Black exonerees.

Race Affects Time Between Conviction and Exoneration

In addition, DPIC’s analysis of exoneration data has found that Black death-row exonerees spend, on average, 4.3 years longer waiting to be exonerated than white exonerees. This translates into more time on death row, more time in prison, and more time after release living under the cloud of a wrongful conviction.

Part of this difference can be attributed to the higher rates of official misconduct and perjured testimony in cases involving Black exonerees — types of intentional misconduct that prosecutors have a greater incentive to hide and charges against which they tend to more aggressively defend. Official misconduct is a factor in 30 of the 33 cases (90.9%) in which exoneration took more than two decades. Perjury or false accusation were present in 28 of those cases (84.8%). But race itself plays a role in delaying exoneration. Even after controlling for the presence of official misconduct, a study by the National Registry of Exonerations still found that race affects how long a person spends in prison. The study concluded that “[i]nnocent African Americans who are convicted of murder are at a disadvantage not only because their convictions were more likely to have been influenced by official misconduct, but also simply because of their race.”5

The racial disparity can be illustrated by who takes the most time to be exonerated. Twelve of the 13 cases in which exoneration from post-Furman wrongful capital convictions took 30 years or longer involved African-American exonerees. Sentenced to death in 1976 under North Carolina’s unconstitutional mandatory death penalty statute, it took 43 years before Charles Ray Finch was exonerated. His wrongful conviction was a product of false forensic testimony and an eyewitness identification manipulated by police misconduct. In March 2019, just months before Finch’s exoneration, the newly created Conviction Integrity Unit in Duval County, Florida freed Clifford Williams, Jr. after 42 years in prison. He had been convicted in 1976 as a result of fabricated eyewitness testimony that was contradicted by the physical evidence. However, his trial counsel failed to contest that testimony and ignored 40 alibi witnesses who placed him elsewhere at the time of the murder.

Time Between Wrongful Capital Convictions and Exonerations Up

Capital cases take a long time to work their way through the courts. That is true both for executions and the even larger number of cases in which defendants come off death row after overturning their convictions or death sentences. But it is also true of exonerations, many of which take even longer because of the failure of courts to detect and acknowledge innocence during the course of ordinary appellate review.

From 1973 through February 15, 2021, death-row exonerees waited an average of 12.0 years from the date of their initial conviction to be exonerated. However, this average includes 17 appellate acquittals based on the insufficiency of the evidence presented at trial, which averaged less than three years each, and more than a dozen cases in which perjury was exposed almost immediately after trial. Death-row exonerees whose cases reached ten years without resolution could expect to wait an average of another decade before they were cleared. That has become the reality for the vast majority of recent death row exonerees. Between 2011 and 2020, as cases tried in the 1990s, 1980s, and even 1970s continued to yield new exonerations, the average time between an exoneree’s initial death sentence and exoneration had lengthened to 22.6 years.

Florida exoneree Clifford Williams and North Carolina exoneree Charles Ray Finch have waited the longest, both being exonerated more than 42 years after their initial wrongful convictions.

Wrongful convictions that take decades to redress are often the result of successful on-going efforts by government officials to suppress misconduct and false testimony or accusations. In all eight of the cases in which an exoneration occurred 31 or more years after conviction, official misconduct was one of the reasons for the wrongful conviction. Official misconduct was also a factor in 88% of the twenty-five cases in which exonerations occurred 21-30 years after the conviction.

Moreover, these exonerations are not examples of a delayed appeal process finally working. Most of these cases escaped detection in the normal course of court proceedings. Leon Brown and Henry McCollum were exonerated in 2014, 30 years after their wrongful convictions in North Carolina, not through the courts, but as a result of an investigation by the North Carolina Innocence Inquiry Commission. Finch was exonerated after he had unsuccessfully exhausted the normal course of appeals, and he was only able to obtain relief because of the pro bono assistance of the Duke Law School Wrongful Convictions Clinic. Williams’ exoneration required the work of the Conviction Review Unit established by State Attorney Melissa Nelson in 2018, which exposed official misconduct by predecessors in the office. Christopher Williams and Walter Ogrod were exonerated in Philadelphia after 26 and 24 years on death row, after the newly created Philadelphia District Attorney’s Conviction Integrity Unit found widespread police and prosecutorial misconduct in their cases.

Pervis Payne’s still-pending case in Shelby County, Tennessee, demonstrates how official misconduct can lengthen proceedings and delay the adjudication of wrongful conviction claims. Payne’s lawyers have argued that Shelby County prosecutors, who have a history of misconduct in capital cases, violated Payne’s rights by withholding exculpatory evidence and making racist arguments to the jury during his trial. Payne was sentenced to death in 1988, but his initial petition for DNA testing was denied in 2006 based on a now-overruled court decision. Prosecutors continued to object to DNA testing in his case, and evidence was finally tested pursuant to a court order in late 2020. In the process of seeking testing, Payne’s lawyers have argued that hidden evidence has been discovered and that crucial evidence such as fingernail scrapings remains missing.

False testimony or accusations have also played a significant role in decades-long wrongful conviction cases. Perjury or false accusation was a factor in 87.5% of cases in which exonerations occurred 31 or more years after the conviction and in 84% of the cases with exonerations 21-30 years after the conviction.

DNA evidence is unlikely to be a factor in exonerations that take place shortly after conviction. It is present in just 7.1% of exonerations that are completed in a less than a decade. However, it is an increasingly prevalent factor in cases that take longer than that to reach resolution. When exonerations occurred more than ten years after conviction, exculpatory DNA evidence is present approximately one-quarter of the time.

Part of this is attributable to prosecutorial resistance to DNA testing, resulting in the need for extensive litigation that delays both the testing itself and the resolution of the case. Moreover, post-conviction DNA testing typically is not available until a defendant has completed his or her initial state court appeal. The first DNA exonerations for death-sentenced prisoners occurred in the 1990s, and DNA played a role in 17-18% of exonerations in the decades between then and 2020. It has been a major factor in the most recent exonerations, with five of the seven exonerations this decade (71.4%) involving DNA evidence.

On the other end of the spectrum, exonerations for insufficient evidence are exclusively seen in exonerations within 1-10 years of convictions. That is because most states require appeals courts to consider the sufficiency of the evidence of a capital conviction on direct appeal, the first stage of appellate review. If the evidence is insufficient to convict, the court need not review any other issue, significantly reducing the time necessary to decide the case. As a result, these exonerations happen more quickly than in the bulk of exoneration cases, which typically require extensive investigation and litigation.

Time Between Conviction and Exoneration by Race

Since official misconduct is more likely to occur in Black exonerees’ cases, the prevalence of official misconduct in long-unremedied cases of wrongful convictions contributes to the longer time Black defendants wait for exoneration. It took Black death-row exonerees an average of 4.3 years longer to be cleared than their white counterparts. African Americans have accounted for 12 of the 13 death-row exonerations that have taken 30 years or longer. Exonerations tend to exhibit a racial preference: two-thirds of white death-row exonerees are freed within 10 years of conviction, while 56% of cases in which Black exonerees who are wrongfully convicted and sentenced to death require more than a decade to resolve.

Race and Exonerations Over Time

Historically, the majority of death-row exonerees have been people of color. Although this general trend has been stable, Latinx representation among exonerees has fluctuated with time and the percentage of Black exonerees has grown in the last decade. Between 1980 and 2010, approximately half of all exonerees each decade were Black, but since 2011 nearly 60% of exonerees have been Black. That may be attributable to the fact that the “new” exonerations are disproportionately cases that have taken longer to resolve and which overwhelmingly involve prosecutorial misconduct. Both of those case characteristics are disproportionately present in cases involving wrongful convictions of Black exonerees.

DNA and Exonerations Up

Most Americans erroneously believe that DNA is responsible for the bulk of exonerations from death row in the United States. That belief is simply a myth. Of the 185 documented death-row exonerations since 1973, DNA evidence has been an exonerating factor in only 28. That translates to 15.1%, or one out of every 6.6 exonerations. By the time Kirk Bloodsworth became the first former death-row prisoner exonerated by DNA in 1993, there had already been 576 other death-row exonerations.

That, however, does not mean that DNA is not increasingly and tremendously important. Beginning with Mr. Bloodsworth’s exoneration, DNA has contributed to 21.9% of death-row exonerations, or 1 in every 4.6. It helped exonerate 8 former death-row prisoners in the 1990s, 10 in the 2000s, 5 in the 2010s, and another 5 in the first 13 months of the 2020s.

The lessons from the DNA evidence are two-fold: it has the power in many cases to prove a person’s innocence. And perhaps equally important, it can show us the unreliability, inaccuracy, or mendacity of the other evidence prosecutors, courts, and juries routinely and erroneously relied upon in convicting innocent people and sending them to death row. For when the DNA proves that a prisoner’s innocence claim is right, it also proves that every other piece of evidence relied upon to convict was wrong.

There is a sense in which the DNA cases are a truth serum for identifying the profile of a potential “innocence case,” for there is no reason to believe that the conduct of police, prosecutors, experts, and informants in cases in which they don’t know DNA will become available will be different from their conduct when it is in fact unavailable. Because there should be no differences between the causes of wrongful convictions in the DNA cases and in the cases with no DNA, the differences that do exist give us important clues about circumstances in which courts should not be denying access to DNA testing and what potentially false evidence they should not be crediting in the absence of DNA evidence.

Yet, while there is no apparent causal relationship between the availability of DNA evidence and the presence of any particular risk factor for wrongful conviction, the differences in the profiles of DNA exonerations and exonerations without DNA are stunning. The presence of every risk factor except perjury or false accusation7 is dramatically different in the DNA-exoneration cases as compared to other exonerations:

  • Official misconduct was a contributing factor in 85.7% of the DNA exonerations (24 of 28), compared to 66.2% in the cases without DNA, a 19.5 percentage-point difference.
  • False or fabricated confessions are present in 46.4% of the DNA exonerations, 4.3 times more frequently and 35.6 percentage points greater than the 10.8% of non-DNA exonerations in which it is present.
  • False or misleading forensic evidence is presented in 71.4% of the DNA cases, 2.9 times the rate and 46.6 percentage points more frequently than the 24.8% of non-DNA exonerations in which it occurs.
  • Inadequate legal defense occurs 35.7% of the time and 1.5 times more frequently in DNA exonerations than it does in the non-DNA exonerations, where its 23.6% level of presence is 12.1 percentage points lower.
  • Mistaken witness identifications occur in 35.7% of the DNA exonerations. That is 2.1 times the 17.2% rate at which it occurs in exonerations without DNA, an 18.5 percentage-point difference.

The numbers strongly suggest that a significant number of additional wrongful convictions involving misconduct, disputed confessions, and questionable forensic or witness testimony have not been detected. The different rates at which these causes appear in DNA versus non-DNA exonerations — especially when there are allegations of more than one of them in an individual case — suggest that they are red flags for DNA testing. Yet, states continue to deny death-row prisoners the opportunity to test DNA evidence that may prove their innocence.

The result may well be wrongful executions — and circumstances suggest that this has already occurred. Posthumous DNA testing indicates that Claude Jones, executed in Texas in 2000, was likely innocent. Tennessee executed Sedley Alley in 2006 after denying him DNA testing that might have established his innocence. The state courts later acknowledged that they misinterpreted their post-conviction DNA-testing law and that Alley’s request for DNA testing should have been granted. After new evidence emerged pointing to a serial killer as the possible perpetrator, Alley’s estate, with the assistance of the Innocence Project, requested posthumous testing. It is still unknown whether the Tennessee courts will allow the testing to proceed.

In Florida — the state with the most death-row exonerations in the nation — a Tampa Bay Times investigation in 2018 found that the state’s courts had refused death-row prisoners access to DNA testing at least seventy times, denying 19 men — eight of whom have been executed — any testing at all and preventing nine others from obtaining testing of additional evidence or more advanced DNA testing after initial tests had been inconclusive.

Courts have denied DNA testing in other cases in which condemned prisoners have presented substantial claims of innocence, including a number of recent cases in which possibly innocent death-row prisoners have been executed.8

The New Exonerations in Context Up

DPIC discovered 11 additional exonerations during its research of all modern-era death sentences. These 11 individuals were tried and sentenced to death in Alabama, Arizona, Georgia, Montana, North Carolina, Ohio, and Texas. Their convictions took place between 1973 and 1993. Nine of the cases involved official misconduct, and five involved both misconduct and perjury or false accusation.

The newly discovered cases share similar characteristics to those in DPIC’s broader list of exonerations.

  • Charles Tolliver, who is Black, was convicted in Ohio by an all-white jury, and after his case was reversed because of the discriminatory use of peremptory strikes, he was acquitted at retrial.
  • John Thomas Alford was jointly tried in North Carolina with a codefendant who had confessed to the murder. At trial, Alford presented evidence and witnesses that he was not at the scene of the crime but the court refused to permit him to present evidence of his codefendant’s confession. He was acquitted after the North Carolina Supreme Court ordered that he be tried separately from his codefendant.
  • The Arizona Supreme Court vacated Andre Minnitt’s conviction because prosecutor Ken Peasley intentionally elicited false testimony from an informant. Peasley was eventually disbarred for his misconduct.

The 11 recently discovered exonerations bring DPIC’s exoneration list from 174 to 185 individuals; however, this number just scratches the surface of the innocence question. DPIC’s list excludes many people in whose cases strong evidence of factual innocence has been uncovered.

Numerous death-row prisoners with strong evidence of innocence have nevertheless taken plea deals to guarantee their immediate release and to avoid the possibility of another wrongful conviction. Prosecutors often offer such plea deals after a reversal to end the proceedings without an official exoneration. In some instances, these deals are offered as a face-saving device by prosecutors who refuse to admit wrongdoing or that mistakes were made, because prosecutors are attempting to preserve the reputation of the office, so prosecutors can escape liability for wrongful prosecution or wrongful imprisonment, or to insulate the state from paying compensation.

After spending 43 years in prison, including 26 on death row, Johnny Lee Gates took a plea deal to secure his release from prison. Gates, who is Black, was sentenced to death by an all-white jury in 1977 for the rape and murder of a white woman. Gates initially confessed to the murder, following what his lawyers charged was coercive police interrogation, and the prosecution withheld key evidence in the case from the defense. Prosecutors claimed that no physical evidence existed in the case until a 2015 Georgia Innocence Project investigation uncovered physical evidence which was DNA tested to exclude Gates. On May 15, 2020, Gates entered an Alford plea, maintaining his innocence but accepting punishment to charges of manslaughter and armed robbery in exchange for a sentence of 20 years on each charge and his immediate release.

Several potential exonerees have died before relief was final. For example, Frederick Thomas’ name is not on the list because he died on death row while Philadelphia prosecutors appealed a court ruling that would have exonerated him. Thomas was convicted on shaky eyewitness testimony and without any physical evidence against him. Before Thomas was granted a new trial in 2002, the state’s two eyewitnesses recanted their testimony and police officer James Ryan—whom the defense said had framed Thomas—was convicted on corruption charges arising out of his conduct in other cases, including falsifying police reports and making false arrests.

In Alabama, death-row prisoners Donnis Musgrove and David Rogers both died on death row before their innocence claims could be adjudicated. Musgrove succumbed to lung cancer on November 25, 2015 while his innocence claim was pending before a federal district court judge. Rogers had predeceased him in 2004.

Rogers’ lawyer, Tommy Nail — who was a state court judge at the time of Musgrove’s death — observed that the case shared “eerie” similarities with that of death-row exoneree Anthony Ray Hinton. The cases were tried by the same prosecutor before the same judge, and the prosecution presented questionable weapons testimony from the same ballistics expert. The ballistics testimony in Hinton’s case was contradicted by three other ballistics experts, and prosecutors decided not to retry him after saying they could not link the bullets from the crime to a gun that belonged to Hinton. Like Hinton, Musgrove and Rogers also had presented solid alibi evidence, Nail said.

In addition to similarly unreliable ballistics testimony, Musgrove’s conviction was tainted by falsified eyewitness testimony, prosecutorial misconduct, and false testimony by a jailhouse informant who later recanted.

Procedural barriers and arbitrary court rulings also may keep individuals with strong innocence claims from being fully exonerated. Tennessee executed Sedley Alley in 2006 after having denied him DNA testing that his lawyers believed could have established his innocence. Five years later, the Tennessee Supreme Court disavowed its decision in Alley’s case, saying they had misapplied Tennessee’s post-conviction DNA testing act. Alley’s family is now seeking DNA testing on behalf of his estate, but prosecutors are contesting the case. They argue that Tennessee law does not allow his family to pursue his case after his death. If prosecutors succeed, a court could block the consideration of DNA evidence that could lead to Alley’s posthumous exoneration.

The addition of eleven more exonerations to DPIC’s Innocence List highlights the systemic failure of the U.S. criminal legal system to protect innocent defendants from wrongful capital convictions and, ultimately, from wrongful execution. The documented failure rate of the judicial process in these cases is alarming: one exoneration for every 8.3 executions. But while DPIC’s exoneration list is an important barometer in assessing the risk of wrongful conviction that innocent capital defendants face, it is a conservative measure of innocence.

Knowing that our nation’s capital punishment system faces a continuing innocence epidemic, it is fair to ask whether capital punishment is worth the risk of imprisoning and executing innocent capital defendants. At what point do the loss of lives or lifetimes to a criminal legal policy that cannot consistently and reliably determine guilt or innocence become too high a price to pay? And when must we say that our state and federal governments cannot be trusted with such a dangerous policy?

This report was updated on March 3, 2021.


  1. They are: Samuel Poole, wrongfully convicted in in Moore County, North Carolina in 1973, exonerated in 1974; Alfred Carey, wrongfully convicted in in Mecklenburg County, North Carolina in 1973, exonerated in 1974; James Creamer, wrongfully convicted in Cobb County, Georgia in 1973, exonerated in 1975; Howard Stack, wrongfully convicted in Fulton County, Georgia in 1973, exonerated in 1975; Christopher Spicer, wrongfully convicted in New Hanover County, North Carolina in 1973, exonerated in 1975; John Alford, wrongfully convicted in Mecklenburg County, North Carolina in 1975, exonerated in 1976; and Thomas Gladish, Richard Greer, Ronald Keine, and Clarence Smith, all wrongfully convicted in Bernalillo County, New Mexico in 1974 and exonerated in 1976. Three of those exonerations, Alford, Carey, and Stack, are among the 11 new exonerations DPIC discovered in during the course of investigating its death-row census.↩︎

  2. https://www.law.umich.edu/special/exoneration/Pages/glossary.aspx. DPIC has adopted the National Registry of Exonerations’ classifications of factors contributing to wrongful convictions.↩︎

  3. The odds that official misconduct contributed to a death-row exoneree’s wrongful capital conviction were 3.7:1 if the exoneree was Black; 2.2:1 if the exoneree was Latinx; and 1.4:1 if the exoneree was white.↩︎

  4. The odds that perjury or false accusation contributed to a death-row exoneree’s wrongful capital conviction were 15:1 if the exoneree was Latinx; 2.4:1 if the exoneree was Black; and 1.4:1 if the exoneree was white.↩︎

  5. http://www.law.umich.edu/special/exoneration/Documents/Race_and_Wrongful_Convictions.pdf; page 7.↩︎

  6. Including 10 of the 11 additional exonerations announced in conjunction with the release of this report.↩︎

  7. Insufficient evidence is not an exception. In those cases, there is an appellate acquittal based on the already developed record in the case. As a result, it never becomes necessary to seek DNA testing.↩︎

  8. E.g., Donnie Lance and Ray Cromartie, Georgia; Ledell Lee, Arkansas.↩︎