On June 29, 1972, the U.S. Supreme Court effectively struck down the death penalty in Furman v. Georgia. 43 years later to the day, in Glossip v. Gross, the justices saved it from one existential threat—but doomed it to another.

Had the Court ruled in favor of three death-row inmates in Glossip, which challenged Oklahoma’s lethal-injection protocol, it would have been a historic shift: Never before has the Supreme Court struck down a state’s chosen method of execution as unconstitutional.

Instead, the justices upheld the use of midazolam, a sedative linked to three botched executions in the past 18 months. “Because capital punishment is constitutional, there must be a constitutional means of carrying it out,” Justice Samuel Alito wrote for a 5-4 majority, ruling that the condemned inmates failed to prove that the drug carried a “substantial risk of harm.” In the principal dissent, Justice Sonia Sotomayor rejected Alito’s logic entirely. “Nothing compels a State to perform an execution,” she argued. “It does not get a constitutional free pass simply because it desires to deliver the ultimate penalty; its ends do not justify any and all means.”

The real debate, however, went far beyond sedatives and protocols. In a surprise move, Justices Stephen Breyer and Ruth Bader Ginsburg filed a separate dissent that called on the Court to revisit the death penalty’s constitutionality.

In 1976, the Court thought that the constitutional infirmities in the death penalty could be healed; the Court in effect delegated significant responsibility to the States to develop procedures that would protect against those constitutional problems. Almost 40 years of studies, surveys, and experience strongly indicate, however, that this effort has failed. Today’s administration of the death penalty involves three fundamental constitutional defects: (1) serious unreliability, (2) arbitrariness in application, and (3) unconscionably long delays that undermine the death penalty’s penological purpose. Perhaps as a result, (4) most places within the United States have abandoned its use.

In the 41 pages that followed, Breyer explained why he believed it was “highly likely” that the death penalty violated the Eighth Amendment. “At the very least, the Court should call for full briefing on the basic question.” Two weeks later, in the first capital case after Glossip, David Zink offered them the means to do so. In his last-minute petition to the Court, Zink presented a question that the Court had not heard in over four decades:

Whether the death penalty today violates evolving standards of decency and concepts of human dignity embodied in the prohibition against cruel and unusual punishment and the Eighth Amendment as applied to the states by the Fourteenth Amendment?

Later that night, the Supreme Court denied Zink’s petition; the state of Missouri executed him at 7:33 p.m. local time. Neither Breyer nor Ginsburg offered any public dissent.

Justices are not required to publicly dissent from cert petitions, although they may choose to do so. Since Furman, three sitting justices have categorically rejected the death penalty on the bench; Lewis Powell and John Paul Stevens also called for its abolition it in retirement. After the Court ended its de facto moratorium in 1976, Justices William Brennan and Thurgood Marshall dissented together in every capital case that ruled against the defendant—1,841 of them, to be precise—until their respective retirements in the 1990s. After publicly announcing he would “no longer tinker with the machinery of death” in 1994, Justice Harry Blackmun also dissented from every capital case thereafter. These acts of judicial civil disobedience are unmatched anywhere else in American jurisprudence. They were also the acts of defeated men. Brennan and Marshall had no hope of pushing the Court back towards abolition; Blackmun admitted in his solitary dissent that he “may not live to see that day, but I have faith that it will eventually arrive.”

Breaking from that tradition could suggest that Breyer and Ginsburg’s silence in Zink was a strategic decision. Evan Mandery, a historian of the death penalty, described the dissent itself as a “Kennedy brief,” wherein a colleague repeatedly cites the current swing justice’s previous opinions in an attempt to elicit his support. “It seems reasonable to surmise that Breyer thinks…that Kennedy’s vote is available, and that the bar should act while the irons are hot in the fire,” Mandery explained. “That, as much as its inherent merits, is why some abolitionists were thrilled by the Breyer brief.”

That doesn’t mean the death penalty’s abolition is inevitable, but its decline is unmistakable. First, the status quo ante SCOTUS doesn’t favor death-penalty supporters. The justices can raise the legal threshold for method-of-execution challenges and deny petitioners’ appeals with ease. But they can neither break the pharmaceutical embargo nor conjure new execution drugs out of thin air.

Second, public opinion also increasingly looks like the pre-Furman landscape. Support for the death penalty is in steady decline across all major demographic groups. 78 percent of Americans backed capital punishment in 1996, according to Pew. That majority shrank to 56 percent in 2015. The sharpest declines since 2011 were among left-leaning groups like women (from 59 to 49 percent) and liberal Democrats (from 40 to 29 percent). But virtually all demographic groups are growing more wary. Support among white mainline Protestants and conservative Republicans fell 7 percent over the last four years; Catholics are now almost evenly split.

Third, and perhaps most interestingly, death sentences from juries are in decline. Executions can be delayed for years by procedural and political factors, and popular sentiment views the situation in abstract terms. But capital juries grapple with the death penalty and those sentenced to it in direct, visceral ways. (Death-qualified juries, as they are known, also exclude committed abolitionists.) In April, I noted how juries in major death-penalty states are sentencing fewer defendants to death with each passing year. That trend continues unabated. In one surprising turn of events, Texas juries didn’t sentence a single defendant to death in the first half of 2015. At the same point in 1998, when death sentences peaked, they handed down 48.

Fourth, the Court has also limited the pool of capital crimes and eligible defendants over the past 25 years. Since reviving the death penalty in 1976, the justices have abolished death sentences for juvenile offenders, for the insane, and for the intellectually disabled. In the 2008 decision Kennedy v. Louisiana, a 5-4 majority forbade the death penalty for crimes “in which the life of the victim was not taken.” In one sense, this trajectory is older than America itself. When the first English legal codes landed in the New World, hundreds of offenses could be punished by death. Colonial-era statutes prescribed capital punishment for over 150 crimes, ranging from rape and murder to “idolatry, witchcraft, and stubbornness in a child.” Now only aggravated homicide and crimes against the state—treason and espionage, for example—remain eligible.

Fifth, many states are still hamstrung when it comes to finding reliable suppliers of execution drugs. Louisiana and Ohio have already suspended executions for the rest of 2015 due to shortages. And despite the Court’s recent sanction of midazolam, no death-penalty state seems eager to adopt it.

None of this preordains the fate of capital punishment, and some of these trends are reversible if the public mood changes. But for now, the cumulative impact of these shifts is fewer capital crimes, fewer death-eligible defendants, fewer death sentences, fewer executions, and fewer viable methods to carry out those executions. In other words, the death penalty is slowly starving.

That doesn’t mean Justice Anthony Kennedy, the swing vote who just finished one constitutional revolution on marriage equality, is willing to lead another one on capital punishment. But the current of events could change that. The Glossip majority missed an opportunity to nudge the states towards more humane methods. Had Kennedy, Roberts, or Alito joined with the four liberals, they could have crafted a narrowly tailored ruling that struck down midazolam and left the overall lethal-injection system intact.

Instead, the majority legitimized the use of unproven drug cocktails by preserving the status quo and erecting new barriers to challenge it. These experiments have already led to botched executions in three separate states. All other things remaining equal, it’s not unreasonable to assume that more will follow. Those deaths already spurred two justices to publicly renounce the death penalty. Would more of them persuade Justice Kennedy, too?