Recently in National polls Category

According to a new Gallup poll, nearly half of all Republicans are non-Hispanic whites who say they attend worship once a week or more. Only one-fifth of Democrats fall into this category and--note this--only a quarter of Independents do. By contrast, 48 percent of Independents, 44 percent of Dems, and 40 percent of Republicans are non-Hispanic whites who attend worship less frequently.

Now, if you're trying to figure out a successful future for the GOP, the best target is those less religious, non-Hispanic white Independents. Independents are swing voters, and that's the biggest bloc of them. They also happen to look like nine-tenths of your existing members. But you're not going to appeal to them by pushing your pro-life, anti-gay marriage agenda. You've got to dial back on the social conservatism. 

What to make of the findings on political identification and church attendance in Gallup's new survey on the Flight from the GOP in the 21st century? The headline is that the only demographic group showing no flight are the frequent church attenders--those who say they go to church at least once a week. What's odd about this finding is that it does not track the shift in this group's voting behavior. By 2000, weekly attenders were preferring Republicans to Democrats in presidential and congressional elections to the tune of 60 percent to 40 percent--the famous God Gap. That 20-point margin held in 2002 and 2004 but in 2006 sunk into the low teens, where it remained in 2008. So if the weekly attenders have voted increasingly Democratic in the past decade, how come there's no sign of it in their party ID?

As this disquisition from Pew makes clear, party ID and voting patterns are different animals, with the former tending to show greater volatility (though, not, apparently, in present instance). It's important to bear in mind that the 60-40 gap among weekly attenders in  2000 translates into a 52-41 gap among Republican identifiers (including leaners) in Gallup's 2000 ID numbers. That's allowing for 11 percent who refused to give even a "leaning" preference. In 2009, the percentage of refusniks went down to 8 percent, meaning that the party ID gap among frequent attenders shrunk by three points, to 52-44.

Still and all, it's clear that the Democrats made significantly less headway among the most observant than they did among the less so. The GOP lost 9 points among those who seldom or never attend and 6 points among the nearly weekly or monthly attenders. Factoring out the non-identifiers, this means that the Democratic advantage among the least observant nearly tripled, jumping 21 points from 51-38 to 63-29; while among the pretty frequent attenders, the Dems have turned a three point deficit (43-46) into a 12 point advantage (52-40).

These latter shifts do track voting patterns, and they suggest two things; first, that the Republicans have driven the least observant voters into the arms of the Democrats; and second, that the Democratic Party's effort to show a more religion-friendly side has borne its fruit among the pretty observant. What's important to recognize is that the latter is the critical religious swing group that Democratic faith-based efforts have always been directed towards.

One way to think of that group is in terms of the abortion issue. As I recently noted, there are a lot of Americans who now call themselves "pro-life" but who support the right to abortion in some instances--according to Gallup's recent poll on the subject, over 20 percent. A large proportion of them are likely to be pretty frequent church attenders. They're the ones who have found themselves increasingly susceptible to Democratic appeals, and it's clear that President Obama will do what he can to keep them in the fold.

Update:  Chip Berlet takes the anti-common ground liberal perspective today over on Religion Dispatches.  A good try, but he's wrong from a practical political perspective, in my view. The Democratic play is to keep the pretty religious from getting scared by culture wars appeals, so they can base their votes on Democratic issues like health care and economic recovery. The left is entitled to resist the temporizing on principle. But it shouldn't kid itself about what's pragmatic.  

The new American Religious Identification Survey is out--for our overall take, see the press release after the jump. The summary report will be is up on its website shortly, but in the meantime, check out  and don't miss USA Today's very cool graphics.

Among the most significant news politically is that 18 percent of Catholics and 39 percent of those belonging to Mainline Protestant denominations say yes when you ask whether they consider themselves born-again or evangelical Christians. That's the same question the exit polls ask, and the one that gives us the results for "the evangelical vote." But we now know that about 25 percent of those "evangelicals" are Catholics and Methodists and Presbyterians etc.--that is, they do not belong to "evangelical" churches. To be sure, mainline churches in the South, Methodist ones especially, are often pretty evangelical. But it's going to take a while to run the cross-tabulations to determine whether these non-evangelical evangelicals are more like other Catholics or mainliners, or more like "true" evangelicals in their beliefs and practices.

Posting on Christianity Today's political blog, Tobin Grant of SIU-Carbondale pooh-poohs the idea that anything significant happened with the evangelical vote for president last November, even as he acknowledges that the exit polls showed a geographical split, with Southern evangelicals more likely to vote for McCain than their Midwestern co-religionists:

The news, however, is that despite the economy, the war, and at least some campaigning by Obama, evangelicals remained unmoved in their support for the Republican candidate.

No, the news from 2008 is the emergence of significant internal divisions within the evangelical vote, not only geographical but also generational. The geographical one (on display in Ted Olsen's cool interactive map) enabled Obama to carry Midwestern states (Ohio, Indiana) that had been beyond the reach of Democratic presidential candidates for a long time. The generational division was portentous, because it showed that among evangelicals, the young went from being the most enthusiastic Bush voters to the least enthusiastic McCain voters, while the old went in the opposite direction. Here's what I wrote about that a couple of months ago:

Laurie Goodstein of the NYT was kind enough to make available some number-crunching of the exit poll numbers on white evangelicals that the pollsters, Edison/Mitofsky, did for her; and it's pretty interesting stuff. The margin among 18-29 year-olds went from 83-16 for Bush in 2004 to 66-32 for McCain in 2008. Among 30-44 year-olds, the shrinkage was from 86-12 to 76-23. Among the 45-64 year-olds, there was essentially no change: 76-23 to 76-22. And among those 65 and older, the GOP margin grew, from 68-32 for Bush to 72-26. So we're talking about swings toward Obama of 33 and 20 points in the younger cohorts, and towards McCain of 1 and 10 points in the older cohorts.

The point, obviously, is that young evangelicals are the future of the voting bloc, and if they hew to their 2008 preferences, the solid 3-1 GOP majorities that evangelicals have turned in for the past few elections is in jeopardy. As with the Catholic vote, aggregate numbers can conceal more than they reveal. 

From the latest WaPo-ABC tracking poll:

Younger white evangelical Protestants (under age 39) break more for McCain than do older people in that group, 85 percent to 13 percent.

We'll see how it all turns out on Tuesday, but I'm betting on white evangelicals splitting not by age but by region.

As for the Jews, the poll has them at 70-29 for Obama, which (if I do say so myself) is what I predicted two months ago:

Revised prediction on the Jewish vote in November: Obama, 70 percent; McCain, 28 percent, Nader 2 percent.

We'll see on that too, eh Kiener?

If you believe the new NYT/CBS poll, then in accounting for his 14-point national lead, Obama has, in only a week, cut McCain's lead among white evangelicals from 55 points (75-20) to 36 (63-27). At that rate, he would pick up a good 30 percent of their vote come November 4--outperforming both Gore and Kerry. On the other hand, the new LAT/Bloomberg poll, which has Obama up by 11, has white evangelicals supporting McCain 69-20, which is more like the usual margin, so who knows?

Quinnipiac's latest on CO, MI, MN, and WI disclose a few salient points on religious voting blocs.
1. In Michigan, where all Catholics barely split for Kerry 50-49, white Catholics are now backing Obama 55-37. Meanwhile, white evangelicals have gone from supporting Bush 2004 76-24, to preferring McCain 58-32. No wonder McCain kissed the state goodbye.
2. White evangelicals in Dobsonland are hard core. In Colorado they went 74-26 for Bush in 2004. Now they back McCain 78-18.
3. White evangelicals in MN and WI go for McCain 65-29 and 57-38 respectively. All told, Midwestern evangelicals do seem to be softening up for the GOP.

Today's Pew poll, which shows Obama's lead growing from two to seven points over two weeks ago, has him losing considerable ground among white Catholics, from seven to 13 points down to McCain. I don't get that. More comprehensible is Obama's turnaround among white Mainline Protestants, from 10 points down to one point up.

Today's Quinnipiac polls of Florida, Ohio, and Pennsylvania show big margins for Obama: 51-43 in Florida, 50-42 in Ohio, and 54-39 in Pennsylvania. He trails by five points among white voters in both Florida and Ohio, but is up by four in Pennsylvania. White Catholics split for McCain 51-44 in Florida, 48-47 and 47-45 in Ohio and Pennsylvania respectively. Nothing much of note there. But the difference between the South and the North when it comes to white evangelicals is striking. McCain leads among them in Florida by the normal (for Republican candidates nationwide) 3-1 margin of 71-24. In Ohio and Pennsylvania, however, it's 2-1: 62-30 and 62-35 respectively. (By contrast, in 2004 white evangelicals in Ohio went for Bush 75-25 in Ohio, 77-23 in Georgia.) This provides some more evidence that, in contrast to the last few election cycles, the white evangelical vote is going to bifurcate--to the benefit of Obama in the swing states north of the Mason-Dixon line.

Latest ABC/WaPo poll shows 47 percent of white Catholics for McCain, 46 percent for Obama.