WASHINGTON (RNS) The largest slice of President Barack Obama's religious coalition -- at 23 percent -- is not very religious.
They're the "nones," also known as unaffiliated voters, according to a new American Values Survey by the Public Religion Research Institute.
Gov. Mitt Romney's biggest bloc of religious voters are white evangelical Protestants, at 37 percent, followed by white mainline Protestants and white Catholics, each at 19 percent. Comparing the candidates' supporters, the more diverse religious and nonreligious coalition that's favoring Obama tends to be younger and growing, which could make it easier for Democrats to win elections in the future.
But there's a down side for Obama, said Dan Cox, PRRI's research director.
"The people most likely to support him are the least likely to vote: Latinos, the millennials (voters 18-29), and the unaffiliated," Cox said.
The study, a comprehensive pre-election survey on the religious and political preferences of American voters, co-authored by the Brookings Institution, zeroes in on the unaffiliated, dividing them into three distinct categories. The report also delves into the Catholic vote, and finds a hardening of the divides within it.
White Catholics today have a stronger preference for Romney than Obama, 54 to 42 percent. That's a wider margin than in 2008, when white Catholics preferred Republican John McCain over Obama, 53 to 47 percent.
With Hispanic Catholics, it's the opposite: a clear preference for Obama, with 71 percent preferring him, compared to 23 percent for Romney.
The survey also highlighted another divide within Catholic voters: those who are more comfortable with their church taking a lead on social justice issues and those who prefer an emphasis on "right to life" issues. About twice as many Catholics (60 percent) fell into the social justice category than the right-to-life category (31 percent.)
Undergirding the various groups' support for candidates is a significant degree of "religious churn in American political life," said Robert P. Jones, PRRI's CEO and founder.
The biggest losers are Catholics. Although nearly one-third (31 percent) of those surveyed said they were raised Catholic, 22 percent currently call themselves Catholic.
On the other end of the spectrum, nearly one in five Americans identify as unaffiliated, but just 7 percent say they were raised that way.
"It's quite remarkable the amount of switching that has gone on," Jones said.
- "Unattached believers" (23 percent) see themselves as religious despite their lack of formal religious ties. They are more likely than the general population to be black or Hispanic and to have completed less education.
- "Seculars" (39 percent) say they are not religious, and reflect most Americans in their racial composition and educational attainment.
- "Atheists or agnostics" (36 percent) are more likely than the general population to be white and highly educated.
The PRRI researchers noted that each of these subgroups tends to have its own views on questions of religion and ethics. Nearly nine in 10 atheists and agnostics, for example, favor same-sex marriage, compared to seven in 10 seculars and nearly six in 10 unattached believers.
These findings on the unaffiliated echo another study, released two weeks ago by the Pew Forum on Religion & Public Life, which also found that atheists are a minority among the nones, most of whom profess to believe in some sort of higher spiritual power.
- About half of Americans say the presidential candidates' religious beliefs are somewhat or very different from their own. For Obama, 49 percent see a disconnect, and for Romney, 53 percent.
- Nearly eight in 10 Romney supporters identify as white Christians, compared to four in 10 Obama supporters.
- There is a big generation gap among white Christians: While seven in 10 seniors put themselves in this category, just three in 10 millennials do.
- Most Americans (56 percent) say religiously affiliated hospitals and colleges should be required to provide health plans that cover contraception and birth control at no cost.
The study of 3,003 Americans, conducted between Sept. 13-30, has a margin or error of plus or minus 2 percentage points.