Can America Kill Its Prisoners Kindly?

A set of lawsuits is pushing the country to find better ways to kill those on death row, with potentially perverse results.

An illustration of a lethal-injection needle with a column of justice
Adam Maida / The Atlantic

About the author: Elizabeth Bruenig is a staff writer at The Atlantic.

Whether killing a person via intravenous poisoning qualifies as cruel and unusual remains, for the moment, an open question. Beginning in late February, the United States District Court for the Western District of Oklahoma heard testimony at the trial of Glossip v. Chandler, an eight-year-old lawsuit filed on behalf of a group of death-row inmates that seeks to prove that Oklahoma’s current lethal-injection recipe—500 milligrams of midazolam, followed by 100 milligrams of vecuronium bromide, followed by 240 mEq potassium chloride—violates the Constitution’s Eighth Amendment ban on cruel and unusual punishment. The case focuses on the constitutionality of midazolam, a sedative with limited anesthetic and no analgesic properties that critics argue results in slow and painful deaths for those poisoned with it.

Recent events have lent the case a special urgency. Last year, John Marion Grant was discovered to have inhaled his own vomit before dying once Oklahoma authorities had injected him with the midazolam-led cocktail, a death that witnesses described as particularly gruesome. Nevertheless, another three executions deploying the same method have gone forward since Grant’s.

The suit will be the latest and most noteworthy in a lengthy lineage of so-called method-of-execution litigation, or legal challenges to capital sentences under the Eighth Amendment. In some states, evidence that lethal injection is inhumane has already led to de facto shutdowns of execution chambers—California, for instance, hasn’t attempted an execution since a sprawling 2006 method-of-execution lawsuit unearthed disturbing facts about the state’s protocol. Glossip probably can’t bank on similar results in Oklahoma, a deep-red state. But that doesn’t mean the lawsuit has no chance of making a difference for Americans on death row.

Success for a suit like Glossip—only the latest development in a body of litigation that has changed capital punishment in America by demanding justice in how the state kills—is a funny thing. Method-of-execution challenges serve an obvious and necessary constitutional function, forcing states to (at least attempt to) keep pace with the “evolving standards of decency” test that the Supreme Court has built into the meaning of the Eighth Amendment. Yet the result of so many decades of method challenges is the lethal-injection regime of the modern era—which, per Glossip, has hardly resolved the inhumane nature of American executions. Therein lies a surreal backwardness: A win for the prisoners in Glossip wouldn’t mean an end to Oklahoman executions, just a ban on midazolam; killings might become marginally more humane, but they also might not—in fact, matters in the future may well be worse. As cases like Glossip multiply and lethal injection becomes a more difficult proposition for states across the board, the legal push for executions that comport with the Eighth Amendment may well result in the resurrection of bygone methods of last resort—the firing squad, electric chair, or gas chamber, with the perverse result that prisoners’ battle for their constitutional rights will, in the hands of the state, become an assault on the very same.

The professor and legal scholar John Bessler writes in his careful history of the Eighth Amendment that America’s ban on cruel and unusual punishment began in part with the case of Titus Oates, an English clergyman who received an excessive (and elaborate) punishment for perjury, making him a cause célèbre among 17th-century British elites and igniting outrage toward King James II. Oates’s aristocratic defenders, inspired by the foment, threw their support behind the English Bill of Rights’ proscription of cruel and unusual punishments, including brutal forms of physical torture as well as exorbitant fines.

Roughly a century later, Bessler told me, the Framers of the American Constitution ratified the ideals set forth in the English Bill of Rights in our own founding documents, in the form of the Eighth Amendment. By the following century, Deborah Denno, a professor at Fordham University School of Law, told me, method-of-execution challenges were already being brought under the amendment’s auspices.

“They started right away with electrocution in the late 1800s,” Denno said. She pointed to the case of William Kemmler, a murderer in New York State condemned to be the first American to be executed by electrocution, in 1890, as the beginning of method challenges in the United States. Kemmler lost his challenge—but, Denno said, his death still marked a turning point in the history of capital punishment. His horrific execution (one witness commented on the “terrible stench” of burning hair and flesh that flooded the chamber as the current coursed through Kemmler) proved the electric chair effective, and attorneys around the country had what they needed to challenge the method in court.

“Those challenges are what start leading us to other methods of execution,” Denno said. The electric chair, a technological marvel, had been introduced to replace the grisly spectacle of hanging—yet it, too, resulted in more than its fair share of needless agony and stomach-churning gore. Tied up in court battles about the legality of their methods, states continued to search for newer, more modern forms of killing—the sort that worked quickly and came with minimal body fluids that might give onlookers the impression that putting a person to death is a somewhat cruel affair.

Lethal injection, legalized in 1977 and first used by the state of Texas to execute a man in 1982, initially seemed to have solved the death penalty’s public-relations problem by introducing a way of killing people with nary a drop of blood nor a wisp of smoke. But, as the Berkeley Law School criminologist Franklin Zimring noted in his 1986 book, Capital Punishment and the American Agenda, within nine years of its development, lethal injection had already collapsed the “fantasy that the intentional taking of human life could be made acceptable to modern standards of decency by changing the means of execution.” In other words, poisoning people to death turned out to be neither as simple nor as painless as its proponents had perhaps imagined.

But even so, the needle gained widespread acceptance, Denno told me, and prosecutors soon found themselves in a bind. “By 2008, just about all the states now had lethal injection,” she told me. “So the rub is, when lethal injection is challenged, [prosecutors] don’t have another method of execution to go to … And by that time, you had a bunch of very sophisticated attorneys who were on to these execution-method arguments.” Denno said that as a result, by the late aughts, “it looked like all these challenges were shutting down states’ abilities to execute anybody.”

While litigation challenging lethal injection’s constitutionality percolated through the courts, those facing execution and their attorneys could indulge in a little uneasy relief: If the method were legally unsettled, then prison authorities were unlikely to move forward, or so it seemed. One method-of-execution challenge, initially filed in 2005 as Roane v. Gonzales, halted federal executions for 15 years as attorneys produced expert testimony and evidence and Department of Justice officials struggled to obtain certain lethal chemicals. One could be forgiven for having believed, just for a little, that the American death penalty was simply dying the way it mandated killing: quietly, privately, paralyzed in the hands of professionals.

The reality is that lethal injection falls short of the replicable accuracy we might expect of medical procedures, and the courts are aware of that. To give one example of a factor that limits the predictability of the method, in a 2019 memorandum, Steven Engel, an assistant attorney general in the Trump administration, concluded that the Food and Drug Administration lacks the authority to regulate chemicals to be used for lethal injections, because the intent is to use them to kill, meaning there’s no way to find them “safe and effective” in the traditional sense. Further, each state sets its own standards for carrying out executions—and fails to meet them in their own ways. A 2006 investigation into California’s death chambers found, for instance, that the lights were kept so dim to protect the identity of prison staff involved with executions that it was unlikely that employees could reliably see what they were doing; that same year, a doctor who supervised lethal injections testified that he was dyslexic and admitted that he sometimes confused the names of medications or the numbers involved in properly dosing them. And accidents do happen: As recently as 2015, an Oklahoma prisoner was killed with the wrong chemical.

Even if there were some way to guarantee that lethal-injection drugs were well regulated and staff were expertly trained in administering them, it would still remain the case that pharmaceutical companies are hesitant to manufacture and sell said chemicals to states for use in executions, thanks to their own ethical concerns and the protests of anti-death-penalty activists. Thus states must still cast about for whatever deadly chemicals they can scrounge up—which leaves the tax-paying public entirely in the dark as to where its money is being spent. In 2017, running short on midazolam, the director of the Arkansas Department of Corrections purchased 40 vials of the chemical for $250 in cash from an unnamed source. In Idaho, correctional officials were accused once again this year of hauling a suitcase full of money to a parking lot for a similar handoff. And in Arizona, authorities spent roughly $1.5 million during a budget squeeze on pentobarbital in unmarked jars to be delivered to an undisclosed location by an unnamed supplier.

Naturally, these irregularities have drawn the attention of journalists, to the profound displeasure of capital-punishment states nationwide. “Since January 2011,” a 2018 report from the Death Penalty Information Center noted, “legislatures in thirteen states have enacted new secrecy statutes that conceal vital information about the execution process. Of the seventeen states that have carried out 246 lethal-injection executions between January 1, 2011 and August 31, 2018 … fourteen states prevented witnesses from seeing at least some part of the execution. Fifteen prevented witnesses from hearing what was happening inside the execution chamber. None of the seventeen allowed witnesses to know when each of the drugs was administered.”

These are all troublesome obstacles to grapple with if one is permitted to serve as a media witness at all. In Oklahoma, only journalists representing local, Oklahoma-based media outlets are permitted to witness executions; only locals are allowed, too, in Texas—where a plurality of U.S. executions occur—save for one slot reserved for the Associated Press. In Alabama, because of the coronavirus pandemic, only a single media witness has been permitted to watch as prison officials administer lethal injections. This means that large, national media outlets have little opportunity to send their own journalists to observe state killings and report on what they see. When I began calling and writing to the Alabama Department of Corrections several weeks in advance of the scheduled execution of Matthew Reeves, who was killed in January, I hoped I might apply to fill the one spot. I never heard back at all.

What don’t these states want you to see? Presumably what I witnessed during the federal execution of 56-year-old Alfred Bourgeois in 2020. Bourgeois’ execution was occasioned by the Trump administration’s secretive procurement of a stockpile of pentobarbital, which Attorney General William Barr directed the Bureau of Prisons to use to execute 13 people in the final months of Donald Trump’s tenure. Several months later, after President Joe Biden had taken office and Attorney General Merrick Garland had issued a moratorium on federal executions, the U.S. District Court for the District of Columbia decided to proceed with discovery in Roane v. Garland, the current iteration of the very same method-of-execution lawsuit filed under Roane v. Gonzales back in 2005. Now the court is considering (among other deaths) whether Bourgeois suffered beyond what this country has chosen to countenance as he was put to death by lethal injection on December 11, 2020, in Indiana.

I was there that evening, only a few yards from Bourgeois. In the chamber with him was a handful of other people, including Jonathan Hemingway, a Federal Bureau of Prisons official who has already submitted his testimony in Roane, on behalf of the government. According to Hemingway’s sworn testimony, Bourgeois never gave any indication of being in any pain or discomfort; he just drifted off into deep breaths and snores before peacefully passing away. Hemingway did briefly see Bourgeois’ “stomach area moving up and down,” though he emphasized a second time that the episode hadn’t made him suspect distress.

I confess that I saw things differently. Like the rest of the media witnesses present, I was stationed in a room separated from the execution chamber by a concrete wall furnished with a one-way window. On the other side of the window was Bourgeois, strapped to a gurney; we witnesses gazed up at his face from below. The angle permitted an excellent view of Bourgeois’ midsection. I watched as he took those effortful breaths with his mouth gaping wide. Then his gut began to churn.

Nobody who witnessed Bourgeois’ death overlooked this sequence of convulsions, though some, as in Hemingway’s case, characterized them as little more than slight changes in posture. In the throes of death, Bourgeois’ midsection undulated with such an eerie, forceful pulse that it put me in mind of looking down at my belly nine months pregnant, when my children were as good as born but not quite and made uneasy peace with captivity by rolling and stretching. But Bourgeois’ spasms looked, even in the moment, like what I suspect they were: his musculature’s effort to draw air into his lungs as he drowned in his own secretions. It was not a quick death.

That should matter. That John Marion Grant vomited repeatedly as he died should matter. In Glossip, it still might: If the suit is successful, midazolam may be found unconstitutional for use in lethal injections and purged from death chambers nationwide. The trial wrapped up earlier this month; a decision is expected as soon as mid-May, and the defense attorneys are hopeful. In such small increments is progress made.

Or, elsewhere, reversed: This year, prison officials in South Carolina announced the completion of a $53,600 overhaul of the state’s execution facilities in preparation for potential killings by firing squad. Like many states, South Carolina has struggled for years to access the scarce lethal-injection drugs available for use. But the state also passed a law last year forcing condemned prisoners to choose an available execution method—either firing squad or the electric chair—should correctional officials prove unable to obtain the chemicals necessary for lethal injection. As of now, South Carolina hasn’t scheduled any executions, by firing squad or another method. But the state is positioning itself to proceed with killings eventually.

Among all the amendments in the Bill of Rights, the Eighth may be the noblest, because it ensures the protection of condemned criminals, the most friendless and vulnerable people. And it should be vindicated, not only for their sake—though mainly so—but also for our own.