When the Supreme Court voided all federal death-penalty laws in June 1972 — despite the Constitution’s clear intent to allow capital punishment — three justices explained that these laws had become "cruel and unusual punishment" because they violated "the evolving standards of decency that mark the progress of a maturing society."

But then, public opinion moved sharply in an unexpected direction. Support for the death penalty soared in the months after the 5-4 decision in Furman v. Georgia. It was 50 percent to 42 percent (an 8-point spread) in March 1972 and 57 percent to 32 percent (a 25-point spread) that November. It kept soaring for two decades, in close sync with crime rates, to a high of 78 percent to 17 percent in the early 1990s. (All numbers are from Gallup polls.)

"Furman, like other landmark cases, had the effect of calling its opponents to action," Stuart Banner wrote in 2002 in The Death Penalty: An American History. One result was the passage of new state death-penalty laws. The justices upheld some of them in 1976 and thereafter because they provided for special hearings to weigh mitigating against aggravating evidence and thus made the process of deciding who dies less capricious (at least in theory). And the number of death sentences pronounced each year by juries rose steadily, to a post-Furman high of 317 in 1996.

But although polls show that about 65 percent of the public still supports capital punishment in the abstract, the number of juries opting for death has plunged, from 317 in 1996 to 128 in 2005, the latest year for which complete data are available. Similarly, the number of executions has dropped from a modern high of 98 in 1999 to 53 in 2006.

At the same time, experts agree, many prosecutors have become more reluctant to seek the death sentence.

And now the Supreme Court has imposed a de facto moratorium on executions while it considers the claims of two Kentucky death-row inmates (Baze v. Rees) and others that the often-botched lethal-injection method used by most states and the federal government may inflict gratuitous pain on condemned prisoners.

The best that death-penalty opponents can hope for in the Kentucky case is a decision requiring states to devise a less error-prone, more pain-free execution procedure. Whatever the outcome, we will probably see a temporary spike in executions after the moratorium ends.

But four factors — more significant than anything that the justices have done or will probably do — seem likely to keep the number of death sentences and executions down in the long run.

•Irrefutable DNA evidence has exonerated some 15 death-row inmates and almost 200 other men convicted of murder or rape, mostly since the late 1990s. This DNA-evidence revolution, along with non-DNA evidence proving the innocence of a great many more condemned men and other prisoners, has alerted many who support the death penalty in principle to the fallibility of the criminal-justice system and the risk of executing innocent people.

•More and more murder defendants have competent trial lawyers, thanks to judicial and legislative decisions requiring more state spending on indigent defense and the work of nonprofits and pro bono lawyers. Few defendants with good trial lawyers get death sentences. And the costs to the state of a well-defended death-penalty trial are often much higher than the costs of imprisoning the defendant for life. All of this has made prosecutors more reluctant to seek death.

•Fewer jurors believe that a death sentence is the only sure way to keep a murderer off the streets. The main reason is that more states — notably including Texas, which leads the nation in executions — have provided life imprisonment without the possibility of parole as an alternative.

(To be sure, a somewhat related reason for the drop in death sentencing — the greatly diminished public fear of crime after the dramatic decline in crime rates between 1994 and 2005 — could prove transitory if the rise in crime rates over the past two years accelerates.)

•Finally, for many centuries people have recoiled against one execution method after another, despite efforts to make them less horrible and less painful. The same seems true now.

Western societies once entertained bloodthirsty crowds with torturous execution methods such as feeding people to lions in the Roman Colosseum and drawing and quartering them. Public executions became progressively less gruesome with the advent of the executioner’s ax, the mechanized guillotine, trapdoor hangings, and firing squads. By the late 19th century, public executions were out of vogue (except for racist mob lynchings) and the supposedly painless electric chair became common. Then came the gas chamber.

And now 36 of the 38 death-penalty states and the federal government have adopted variants of the three-drug lethal-injection cocktail that is before the Supreme Court: first a sedative to put the prisoner asleep; then a paralyzing agent to prevent struggle; then the drug that kills by stopping the heart.

But like all previous methods, the three-drug protocol has often been botched, in ways dramatically detailed by defense briefs in the pending Supreme Court case. This is unsurprising: Not many physicians — indeed, not many people with any kind of marketable skill — aspire to be executioners.

Does this pattern suggest that our raw retributive urges recede as we become more civilized? How many death-penalty supporters would choose to play even the most clinical role in an execution? How many Americans would relish the spectacle of murderers (let alone adulteresses) being publicly stoned to death, as they still are in some of the same societies that spawn barbarian suicide bombers?

On the other hand, what explains the insistence of so many state governments on keeping a hastily devised, demonstrably flawed lethal-injection protocol that may well cause agonizing pain? It would not be difficult to kill prisoners painlessly. A bullet to the back of the head would do. But who would volunteer to pull the trigger? A massive dose of the barbiturates used to put down sick pets would also suffice. But might that seem too dehumanizing? Or, perhaps, less awe-inspiring than an execution should be? Is there a good way to do this? If not, should we be doing it at all?

For these and other reasons, I suspect that the abolitionist justices may have been right in their perception that the death penalty is in tension with the "evolving standards of decency that mark the progress of a maturing society." (Chief Justice Earl Warren originated the line in 1958, in Trop v. Dulles.)

They were wrong, however, to take this perception as a mandate to abolish the death penalty, for at least three reasons.

First, the recent history sketched previously shows that the trend toward less enthusiasm for executions is not linear, and that well-intentioned judicial efforts to speed up the evolution process can backfire.

Second, judicial abolition of the death penalty would usurp powers assigned by the Constitution to the legislative and executive branches. In this sense, Furman had less to do with the progress of a maturing society than with what Justice William Brennan, when breaking in new law clerks, introduced semi-facetiously as "the rule of five."

"You got five votes," he would explain with mock professorial gravity, "you can do anything you want." So you can, for a while. But in the long run, societies mature better when they do it the old-fashioned way, without the help of judicial fiats.

Third, while the justices know a lot about the social costs of the death penalty, they know very little about whether and to what extent these costs may be offset by the very considerable benefit of saving innocent lives.

Pepperdine University professors Roy Adler and Michael Summers, for example, claim in a November 2 Wall Street Journal op-ed that "our recent research shows that each execution carried out is correlated with about 74 fewer murders the following year." If you believed that, wouldn’t you want to see a lot more executions? Indeed, at least in the case of rational criminals, any deterrent effect would presumably increase in tandem with the (now tiny) risk of being executed.

Many more experts argue that the death penalty does not deter crime. And it seems likely that in some contexts — jihadist terrorists who crave martyrdom, for example — more executions might actually bring more murders.

Putting aside such special cases, at this point most of us can only speculate about which side has the better of the inherently conjectural arguments about deterrence. The same is true of the justices. And speculation is not a firm foundation on which to build constitutional law.