The belief that we can tell the guilty from the innocent and those deserving of death from those deserving of life has many roots, but one of the most powerful may be the use of science—or what purports to be science— in criminal trials. Scientific evidence often comes to court with an aura of infallibility, appearing to offer jurors certainty that they will not convict (or sentence to death) an undeserving person.
But while there is no doubt science has much to offer the criminal system’s core truth and justice-seeking missions, our experience at the Innocence Project demonstrates that when purported scientific evidence is not validated or reliable, it has a devastating effect on those same aims. Indeed, the convictions in almost half of the 329 DNA exonerations rested in part on unvalidated and unreliable scientific evidence. Thirteen of these innocent people, like Ray Krone and Dennis Williams, were also sentenced to death. These are not isolated problems. Just last month, The Washington Post reported that “[t]he Justice Department and FBI have formally acknowledged that nearly every [hair microscopy] examiner in an elite FBI forensic unit gave flawed testimony in almost all trials in which they offered evidence against criminal defendants over more than a two-decade period before 2000.”
Unvalidated and unreliable forensic science is being used not just to determine who will die, but how. On April 29th, the Supreme Court heard oral arguments in Glossip v. Gross, a case that challenges Oklahoma’s use of an anti-anxiety drug called midazolam in its lethal injection procedures. Oklahoma, like many states, uses a three-drug protocol in executions. The first of these drugs is intended to “induce a deep, comalike unconsciousness,” before two other drugs are used to induce paralysis and death. This deep unconsciousness is necessary to ensure those injected with the “liquid fire” of the third, killing drug do not feel constitutionally intolerable pain.
After European manufacturers took measures to prevent their products from being used in executions, Oklahoma turned to midazolam to create this unconscious state. In 2014, Oklahoma used midazolam for the first time in the execution of Clayton Lockett. Even after receiving all three drugs, including 100 milligrams of midazolam, and being declared unconscious, Lockett began to “writhe and gasp,” suggesting he could feel the “agonizing suffocation and pain” caused by the two other drugs. (Two other men, Joseph Wood in Arizona and Dennis McGuire in Ohio, suffered similar fates during their executions.) After Lockett’s botched execution, Oklahoma increased the midazolam dosage to 500 milligrams.
Glossip and a group of other petitioners challenged this protocol in federal court. At an evidentiary hearing, Oklahoma put on only one witness, Dr. R. Lee Evans, to support its use of the drug. Evans testified that 500 milligrams of midazolam would render a person sufficiently unconscious for purposes of execution. But this conclusion has no basis in science, and no reliable scientific data or methodologies supported it. Rather, Evans merely presumed that because midazolam would kill a person at a specific dose, that same dose would also cause a sustained coma before causing death—a conclusion belied by the many botched midazolam executions described above. As. Evans himself admitted, only “extrapolation” and “assumption” supported this opinion. Indeed, because no peer-reviewed scientific literature—one of the hallmarks of validated and reliable science—supported his extrapolations, Evans relied instead on the website drugs.com. In addition, Evans made a material miscalculation in determining the drug’s toxic dose in this first instance. It was on this unscientific testimony that the district court, and ultimately the appeals court, upheld Oklahoma’s use of the drug.
Unsupported speculation and mathematical errors are not science. Science requires a rigorous application of the scientific method, which in turn demands empirical testing methods, peer review, and objective standards. Whether the Court will ultimately acknowledge that purportedly scientific testimony which falls short of science also falls short of what is constitutionally necessary to take a person’s life remains to be seen. At oral argument, several of the Justices, Justice Breyer in particular, evinced serious concern about the lack of scientific basis for Dr. Evans’s opinion. Whatever the Court decides, it is plain that basic notions of fairness and justice require that only validated and reliable forensic science be used to determine guilt and punishment; no less do they compel us when determining how such punishments are carried out.
Dana Delger is a staff attorney in the Strategic Litigation Unit of the Innocence Project.