For 40 years American politicians have assumed that favoring the death penalty is a winning political position. Is that era coming to an end? Is support for capital punishment, like opposition to gay marriage, evaporating? We can't be sure. But we're seeing the first signs that it could happen.
Death penalty support peaked at 80 percent in 1994 in the Gallup poll and the National Opinion Research Center's General Social Survey. Since then, it has been sliding. In the most recently published GSS sample, taken in 2012, support fell to 65 percent, the lowest number since the question was introduced in its current form four decades ago. If it falls any further, it'll be in new territory. The latest Gallup sample, taken last year, found that support was down to 60 percent for the first time in 40 years.
In a Pew survey taken a year ago, support for executing murderers dropped to 55 percent, 3 points down from Pew's previous low. Last month, in a CBS News survey, the support level fell to 59 percent (4 points down from the previous low) while the percentage of respondents who opposed the death penalty rose to 33 percent (6 points above the previous high). It's the first time in the 26 years CBS News has asked this question that the support number has fallen into the 50s or the opposition number has climbed into the 30s.
A Washington Post/ABC News poll released this month points in the same direction. Given a choice between two punishments for murder, only 42 percent chose the death penalty. Fifty-two percent preferred life imprisonment without parole. That's an 8-point drop in support for capital punishment since the previous Post/ABC poll in 2006. It's the first time in recent history a majority has chosen life over death.
Why is enthusiasm for the death penalty declining? Will it keep falling? Let's look at what has changed.
Crime rates. Academic analyses begin with this correlation. The rise and fall of death penalty support since 1960 closely tracks the rate of homicide and other violent crimes. If crime rates continue to fall, capital punishment could sink with them.
But for the same reason, if crime increases, support for the death penalty could rise with it.
Deterrence. In Gallup's 1985 and 1986 surveys, respondents agreed by roughly 2-to-1 ratios (61 percent to 32 percent in 1985, 62 percent to 31 percent a year later), that capital punishment "lowers the murder rate." By 1991 the percentages had shifted by about 10 points. By the 2000s, the 2-to-1 ratio had completely reversed: More than 60 percent rejected the deterrence claim. That's a 30-point swing in 20 years. Harris polls show a similar trend. From the early 1980s to the 2000s, the percentage of respondents who believed that executions deterred murder fell nearly 20 points.
This is an empirical belief, not a moral one. There is an academic debate over whether executions affect the murder rate. The question is difficult to resolve in part because the number of executions is too small to provide a clear answer.
Life without parole. Preference for this punishment, as an alternative to execution, increased in Gallup polls from 1985 to 2010. In the latest Washington Post/ABC News survey, even in states that have the death penalty, a plurality preferred life without parole. But polls show that overall opposition to the death penalty has been growing as well. In Gallup's trend data, the change in death penalty support (20 points) exceeds the change in response to the life-without-parole question (16 points).
In 2007, when the Death Penalty Information Center took a poll to find out why people were turning against capital punishment, the availability of life without parole didn't seem to make much difference.
Execution methods. Recent lows in some polls have followed media reports of botched executions. But botched executions have been reported for decades, apparently without sustained effects on public opinion. The GSS, Pew and Gallup lows took place during a multiyear lull between such reports. When the Post/ABC poll presented a scenario in which lethal injection was "outlawed or otherwise unavailable," 10 percent of respondents shifted from supporting the death penalty to saying it should end. But in response to a similar question, presented in an NBC News poll last month, 61 percent of respondents chose an alternative method - more than the 59 percent who originally said they favored capital punishment.
Moral beliefs about killing. In 1991 and 2003, Gallup asked people why they opposed the death penalty. Between the two polls, opposition to the death penalty rose from 18 percent to 28 percent, and the percentage of death penalty opponents who cited the wrongness of taking life (each respondent could volunteer just one reason) rose from 41 percent to 46 percent. That works out, by multiplication, to an increase of 5 or 6 points in the percentage of respondents overall who opposed capital punishment and said their principal reason was that killing is wrong - accounting for about half the increase in death penalty opposition.
I'm skeptical. In these polls, the margin of error is 5 percent for the entire sample and goes up as you work with subsamples. If you take seriously a 5-point increase in the percentage of death penalty opponents who cited the wrongness of taking life, what about the 4-point decline in the percentage of death penalty opponents who cited religion? Isn't that a wash? Maybe people who thought it was wrong to kill just stopped mentioning God.
In 2001 Gallup added a question that could detect a change in ethics. For capital punishment and other issues, the question asked: "Regardless of whether or not you think it should be legal ... please tell me whether you personally believe that in general it is morally acceptable or morally wrong." In seven polls taken from 2001 to 2007, on average, 66 percent of respondents said the death penalty was acceptable; 27 percent said it was wrong. In seven polls taken from 2008 to 2014, the "acceptable" average fell to 62 percent, and the "wrong" average increased to 30 percent.
Are people becoming more averse to violence? Over the long term, yes, according to evidence presented by Steven Pinker. But it's hard to connect that grand arc with public opinion trends on the death penalty.
Innocence. In Gallup's 1991 and 2003 surveys, when death penalty opponents were asked for their reasons, the biggest shift by far was in the percentage who cited the risk of erroneous conviction. That number more than doubled, from 11 percent to 25 percent of the anti-death-penalty subsample. This finding is backed up by the Death Penalty Information Center's 2007 survey, which identified people who had shifted from supporting to opposing capital punishment. When they were asked which of several factors influenced them most, 62 percent cited "evidence that innocent people are sometimes sentenced to death."
In last month's NBC News poll, the most popular reason to oppose capital punishment - cited by 35 percent of respondents - was that "it carries the risk of killing someone who was wrongly convicted."
But there's a catch. In the same NBC News poll, the second-most-popular reason given for supporting the death penalty (and the best reason to support it, according to respondents who themselves oppose capital punishment) was that "modern science, like DNA testing, reduces the possibility someone has been wrongly convicted." Apparently, many people are willing to believe that science, having exposed the problem of wrongful conviction, can fix it. If so, those who have turned against the death penalty to avert lethal error could regain their confidence and reinstate their support.
We got used to high levels of public support for capital punishment in the 1980s and 1990s. But it wasn't always so: The death penalty was far less favored in the 1950s and 1960s, sometimes lacking majority or even plurality support. The lows we're seeing today don't guarantee a further slide - this century's numbers could bounce around as much as last century's.
But when you look at the array of surveys descending into unfamiliar territory, and when you study the factors behind this descent, it's reasonable to think it could keep right on going.
Saletan (@saletan) covers science, technology and politics for Slate.