Why is it so hard to correct mistakes on death row? 

Robert Roberson sits on death row in Texas for a crime he maintains he did not commit. Convicted and sentenced in 2003 on charges of killing his two-year-old daughter, Roberson is scheduled to be executed on Oct. 17. 

Late last month, the man who served as the lead detective in Roberson’s case told the New York Times that there is what he called “unassailable doubt” about Roberson’s guilt, that the evidence against him is no longer believable and that it needs to be re-examined. He said, “I’m convinced we did the wrong thing,” and called what happened to Roberson “our failure.” 

That is one of the many reasons his execution should be stopped and why Roberson should not be in prison at all. But, at this point, it is unclear whether any court, or Texas Gov. Greg Abbott, will rectify that miscarriage of justice. 

While there have been at least 200 death row exonerations in the last 50 years, not one of them has come easily. Each represents an enormous success and tribute to the extraordinary work done by lawyers and advocacy groups across the country.  

Their work also has helped change the way Americans think about capital punishment; it helps explain why more Americans now think that the death penalty is administered unfairly than fairly. 

But the fact that there are so many roadblocks erected to prevent success in freeing falsely convicted inmates from death row, and that it takes so long to bring about an exoneration, is truly scandalous. While we celebrate the exoneration movement’s success, we must work to address and remove those roadblocks. 

This will require re-examining the procedural hurdles that make it difficult for courts to consider evidence of innocence uncovered long after someone has been convicted of a capital crime. It will also require addressing the attitudes of prosecutors and law enforcement officials who often seem to care more about defending a conviction than about getting to the truth, or acknowledging when a miscarriage of justice has occurred. 

A new report by the Death Penalty Information Center highlights the price paid by the victims of those miscarriages of justice. Despite the success that death penalty opponents have had in raising awareness of the ubiquity of false convictions in death cases, “in the past twenty years, the average length of time before exoneration has roughly tripled, and 2024 has the highest-ever average wait before exoneration, at 38.7 years.” 

It is hard to imagine the horror of being convicted and sentenced to even one day in prison for a crime you didn’t commit. Compound that horror every day for over 38 years, and you get a sense of what is so scandalous about America’s many miscarriages of justice and the difficulty of rectifying them when they occur. 

As troubling as that is, Roberson’s case reminds us that many other death row inmates have plausible claims that they were falsely convicted but for whom exoneration remains a distant prospect. 

In his case, as the Innocence Project reports, “Nikki, Roberson’s chronically ill daughter, was sick with a high fever and suffered a short fall from bed” before her death. Her illnesses had been misdiagnosed by doctors who prescribed medication that should not have been given to such a young child.  

At the time of her death, doctors ignored those factors, suspecting instead that she was a victim of what was then called Shaken Baby Syndrome.  

The police quickly concluded that Roberson was the culprit.  

As the Innocence Project notes, “Roberson’s own defense lawyer agreed with the State that Nikki must have died from SBS. When Roberson refused to accept a plea deal, his lawyer argued only that Roberson had not meant to kill Nikki and that he was mentally impaired.” 

Shaken baby syndrome was first recognized in the 1970s; in 2001, two years before Roberson’s trial, the American Academy of Pediatrics “said that where a young child had internal head injuries there was ‘the need for a presumption of child abuse’ — a categoric finding that rapidly became medical, and then forensic dogma.” 

Today, however, the science of Shaken Baby Syndrome is under question. Since Roberson’s trial, it has been discredited and identified by some as junk science. Last year, a New Jersey Appeals Court labelled Shaken Baby Syndrome “scientifically unreliable” and said it could not be used by prosecutors in trials in that state.

Roberson is not the only person falsely convicted of a crime based on that particular example of junk science. In June, The Innocence Project observed that “To date, at least 32 parents and caregivers in 18 states have been exonerated after being wrongfully convicted under the shaken baby hypothesis.” 

Moreover, in Roberson’s case, as the Death Penalty Information Center explains, “The jury was unaware that Nikki had pneumonia at the time of her death and that she had been prescribed medication — no longer recommended for children her age — that can suppress breathing. CAT scans also showed a single impact injury to her head, consistent with impact from a fall. Nikki had no injuries to her neck, which one would expect had she been subjected to violent shaking.” 

And, compounding the scandal surrounding that case, in 2013, the Texas legislature passed a law that allows prisoners to challenge their wrongful convictions “by showing that changes in the field of forensic science either undermined the integrity of the criminal trials that led to their convictions or exonerated the defendant.”  

However, that statute has not helped Roberson. A study released by the Texas Defender Service this month explains why. 

It found that “the Texas Court of Criminal Appeals’ implementation of the statute has shown inconsistency in application, a disregard for discredited scientific methods, a heavy investigative burden for people seeking relief (especially people without counsel), and a striking absence of relief in capital cases — meaning that potentially innocent people will be executed.” 

If Texas gets its way, Roberson will be one of them.  

So far, the courts have not seemed to care. Last year, he sought relief from the United States Supreme Court, alleging that due process is violated “when the entire causation theory of the crime was undermined by modern science and medicine, and when the habeas court ruled against the defendant based on outdated forensic theories.”  

In October, the court refused to hear his case. So Roberson remains on death row, hoping for vindication of his innocence claims before it is too late.  

He is not alone. 

The new Death Penalty Information Center report points to what it calls “longstanding efforts to restrict the number and scope of appeals.” It highlights the Anti-Terrorism and Effective Death Penalty Act that “imposed a strict one-year deadline for federal post-conviction appeals and narrowed the grounds upon which federal courts could grant post-conviction relief.” 

Yet in cases of wrongful convictions brought about by “official misconduct, false witness testimony, improper forensic evidence or ‘junk’ science, and inadequate legal representation …evidence to support those claims is sometimes slow to emerge by no fault of the prisoner.”  Consequently, “many prisoners with credible innocence claims are effectively boxed out of court due to draconian procedural rules.”  

In addition, unlike the chief investigator in the Roberson case, prosecutors are often reluctant to acknowledge errors and expose the official misconduct that is often responsible for wrongful convictions.  

As the clock ticks toward Roberson’s execution, the need for action is urgent. But so is the work that remains to be done to undo the scandalous delays and barriers to getting the people off death row who do not belong there. 

Austin Sarat is the William Nelson Cromwell Professor of Jurisprudence and Political Science at Amherst College. His views do not necessarily reflect those of Amherst College.