After the Donald Trump administration revived the long-dormant federal death penalty, the Biden administration appears to be putting it back to sleep. The Justice Department has issued a moratorium on federal executions, and federal prosecutors have quietly dropped capital charges in seven recent cases.

This is consistent with what presidential candidate Joe Biden promised, but incomplete. His campaign website stated, “Because we cannot ensure we get death penalty cases right every time, Biden will work to pass legislation to eliminate the death penalty at the federal level, and incentivize states to follow the federal government’s example.”

But no legislation to eliminate the death penalty has been submitted, and we have heard little from the White House about those incentives to the states. Meanwhile, the administration has opposed requests from two death row inmates, Boston Marathon bomber Dzhokhar Tsarnaev and Charleston church killer Dylann Roof, to be resentenced to life without parole.

If President Biden wanted, he could halt the federal death penalty for a generation with the stroke of a pen by commuting the sentences of the 46 people now on federal death row.

If the president and his advisers think this would be unpopular, they are wrong. By analyzing 595 public opinion surveys, I have found that Americans support capital punishment less than they have at any time since the modern death penalty system was established in 1976.

What 595 surveys tell us

I recently updated a comprehensive review of every nationally representative public opinion poll on the topic of capital punishment since 1936, building on research conducted with co-authors in a 2008 book, and with others in 2018.

I studied polls beginning with a 1936 Gallup poll, in which they asked whether respondents believed in the death penalty for murder and found that 61 percent said yes, to another Gallup poll May 3, 2021, in which 55 percent said they believed the death penalty was morally acceptable and 40 percent said it was morally wrong. Drawing on the Roper Center iPoll database, I found 595 survey estimates in the 85 years between.

Many of the polls worded the questions differently — and as in all polling, question wording matters a lot. Furthermore, in the 595 surveys, 147 used specific question wording only once, in some cases, asking about specific offenders; that wasn’t useful for comparing attitudes over time. However, in 448 of the polls, the same survey organization posed the same question to a national survey of adult Americans more than once.

Across these 448 polls, there were 65 differently worded questions. The most common was the familiar Gallup poll question, posed 57 times: “Do you favor or oppose the death penalty for persons convicted of murder?” The following figure shows how Americans have answered that since it was first asked in 1953; as you can see, in recent years, support has dropped and opposition has grown.

Combining several survey results

However, this question’s wording may influence the responses. For example, the figure below compares support for the death penalty in response to two different Gallup poll questions: First is the one just described: “Do you favor or oppose the death penalty for persons convicted of murder?” and second is another common way of asking the question: “What do you think should be the penalty for murder: The death penalty, or life imprisonment with absolutely no possibility of parole?” When given the opportunity to choose life without parole, fewer people support the death penalty — on average, support is 12 points lower.

To address this kind of variation, my University of North Carolina political science colleague Jim Stimson developed a mathematical technique and the software to calculate trends from several different survey questions on the same topic, even if these surveys word the question in different ways. This allows us to see whether the trend of public opinion on the death penalty is moving in one direction or another, although the different wordings mean that we can no longer interpret the “percent agreeing” with the question as the share of Americans who support capital punishment.

A single comprehensive summary of the state of public opinion

The following figure pulls together the results from every bit of usable public opinion evidence available since 1953.

If we average attitudes toward the death penalty from 1953 to 2020, we find that in 1953, support was just above average. It then fell to its low in 1966 before trending unevenly but steadily upward to its high in 1997, with support for capital punishment nearly 20 points higher than in 1966. But since then, support has declined again to its lowest point since 1966.

In other words, public opinion toward the death penalty has soured even more quickly in the past two decades than it grew in the decades before. In the late 1990s and early 2000s, as innocence projects began using DNA tests and other previously unavailable tools to prove that many people on death row had not committed the crimes for which they’d been convicted, Americans began to see that police, prosecutors, judges and juries are imperfect and can sometimes make mistakes.

Many Americans, both Democrats and Republicans, were troubled by the idea that innocent people were being executed, leading to a dramatic and rapid decline in support for the death penalty.

What this means for the president and the attorney general

Death sentences have steadily declined since 1996, when 315 people were sentenced with death, to 18 people in 2020. Executions too have declined (from 98 in 1999 to 17 in 2020). Most of those were carried out by state governments. The federal government executed only three individuals since the 1972 Furman decision until the last months of the Trump administration, when it executed 13 individuals.

As the Biden administration considers its next steps, officials will probably pay attention to public opinion, which now opposes the death penalty more than at any time in the previous 50 years.

Frank R. Baumgartner is the Richard J. Richardson distinguished professor of political science at the University of North Carolina-Chapel Hill.