Religion and Presidential Voting Patterns

My friend Henry Olsen knows election data just about as well as anyone in the public eye. Since he now has a brand new blog and a forthcoming book to promote, I figured I'd take the opportunity to ask him a few questions about the prospects for religion in American electoral politics. My questions and his answers appear below.

Henry argues that there are four factions among GOP voters: moderates, “somewhat conservatives,” religious conservatives and secular conservatives. Presidential nominations tend to go to the candidate who can unite the “somewhat conservatives” with one of the other factions.

The data and observations below suggest that the substantial divide in the GOP between us religious conservatives and other kinds of conservatives is about something much bigger than policy; they just don't share our view that the moral center of the nation is disintegrating. If the most noteworthy policy difference between us is whether abortion bans should make exceptions for rape, I'd say we're not far apart on policy. But the somewhat conservatives who have the most weight at the GOP negotiating table simply don't think, as I think most of us in the religious conservative group do, that our civilization is faced with a crisis of ultimate commitments. They want “stature and gravitas,” a reassuring figure who promises that our problems can be fixed if we just put the grown-ups back in charge in Washington, without the need for a renewed commitment on our own part to the hard virtues like self-control, frugality and justice. (Note Henry's casual reference to “fringe candidates who stress moral decline.”)

There are good philosophical reasons not to seek a general moral restoration of the public square through electoral politics, but on top of all that, the votes aren't there and they're not going to be there. We must continue to fight for justice in the public square, but we must realize that the voting booth is the rearguard, not the vanguard, of that fight. There are other ways to support justice in our political system besides elections—voting the right people into office—and there are other aspects of the public square besides the political. It is there—almost certainly in cooperation with the “Faith and Family Left”—that we ought to turn our primary attention.

As that (alas) incorrigible atheist Milton Friedman used to say, “don't change the players, change the game.”

Q: To what extent do the religious and secular conservative factions differ over actual positions (e.g. pro-choice secularists, evangelicals against free trade), to what extent do they agree about positions and differ over issue priorities, and to what extent are the differences between them more symbolic or sociological rather than policy differences per se?

Henry: The exit poll data are limited, so I can only give you provisional answers. However, on abortion very conservative evangelicals (or, when asked, frequent church attenders) are much likelier to oppose abortion in all cases than any other group of conservatives. Other groups of conservatives are likelier to back restricting abortion in most cases, and somewhat conservatives have a higher share of people who are some degree of pro-choice. Very conservative evangelicals are likelier to back fringe candidates who stress moral decline and social issues than are other conservative factions. In the 2000 Iowa caucus, for example, over 40 percent of very conservative evangelicals voted for either Alan Keyes or Gary Bauer even though co-religionist George W. Bush was the prohibitive favorite for the nomination.

Very conservative evangelicals also say social issues and religiosity are more important in their voting decision than do other conservative factions. In 2012, 20 percent listed abortion as the most important issue they considered, triple the number who said this among other Republicans. They are also much likelier to say that values, especially conservative values, are more important than winning or experience when considering whether to support a candidate. Finally, they very much want to know a person's religious beliefs. In 2008 and 2012, nearly 90 percent said the fact that a candidate shared their religious values mattered “a great deal” or “somewhat” when deciding who to vote for, about double the share of all other Republican voters.

Q: If your theory is right, the disposition of the “somewhat conservatives” will be very important to the near future of the GOP. Do you have a read on how religious or secular they are, and to what extent (if at all) this might influence which of the two fully conservative factions they ally with?

Henry: The exit poll data do a poor job of describing the “somewhat conservative” voter, in part because they do a good job in dividing right from left. On virtually every issue, the somewhat conservative voter sits between the moderates and the very conservatives. They are more likely than very conservative evangelicals to highly value a candidate's experience or their ability to beat the Democrat, and less likely to say sharing a candidate's religion or values is of prime concern. This extends to issues: fewer than ten percent say that a candidate's stance on abortion matters most to them. They are, nevertheless, more likely to be pro-life than pro-choice, although only one in seven say abortion should be outlawed in all cases.

We can see much more about them by looking at who they back in early and later races. They always back the man with stature and gravitas, at least compared to his rivals. Dole in 1996, Bush in 2000, McCain in 2008, Romney in 2012: firebrands need not apply to the somewhat conservative tent.

In one-on-one races, very conservative seculars tend to line up behind the somewhat conservative favorite rather than vote for the very conservative evangelical candidate. This is why Santorum did not gain much after Gingrich suspended his campaign: Romney gained more votes from former Newt backers than did Santorum, especially so among somewhat conservatives.

Q: Do religion and secularism have more voter salience in the GOP than among the Democrats? What observations do you have on the role of religion in the Democratic party?

Henry: I have not studied the Democratic Party primary electorate in great detail, so I cannot really give you a good answer. I can point to data from Pew Research which breaks the entire electorate into eight separate groups. “Solid Liberals,” the most partisan Democratic group, is easily the least religious group followed closely by the “Next Gen Left”. Only 19 percent of the former and 21 percent of the latter attend religious services weekly or more. 55 percent of “Steadfast Conservatives” attend services that regularly.

One Democratic group, however, does look more like Republicans on religion: the “Faith and Family Left.”  51 percent attend services at least weekly, and they are the group least likely to say they never or seldom attend. They are also second (behind Steadfast Conservatives) to reject evolution and to agree with the statement “you religion's holy book is the word of God.”  Solid Liberals, in contrast, are the group likeliest to reject that view in favor of the idea that “your religion's holy book is written by man not God.”

The Faith and Family Left is the least white of all eight groups. Fully thirty percent are black (twelve percent are black nationally) and another nineteen percent are Hispanic (versus thirteen percent nationally).

Data not cited to the Pew Foundation are from my forthcoming book, The Four Faces of the Republican Party: The Fight for the 2016 Presidential Nomination (Palgrave, November 2015).