Is it reasonable to think that Barack Obama can woo enough evangelicals away from the Republican Party to pull off victories in some of the crucial battleground states?
The answer might depend on which polls you are looking at.
Evangelicals have been recognized as pivotal players in the American political landscape at least since the election of President Jimmy Carter. As a vital component of the values voter bloc, along with conservative Roman Catholics, they are even credited with making the crucial difference in the 2004 presidential race.
But assessing the precise role of evangelicals in elections is not easy, a new study by two sociologists shows, because analysts and pollsters use quite different definitions to identify these voters and their movement. Noting that claims about the size and characteristics of American evangelicals are "often inconsistent, even contradictory," Michael Lindsay, a professor at Rice University, and Conrad Hackett, a postdoctoral fellow at the University of Texas, say that estimates of the adult evangelical population have ranged from as small as 7 percent of the total adult population to as large as 47 percent.
And that, as they say in the gambling world, is a pretty big point spread.
So should we simply ignore all that polling data about evangelicals? No, the two authors say, but we should be aware of how different types of polls go about studying them.
To clarify matters, Lindsay and Hackett explain the three distinct ways in which leading national polls do so. As used in the General Social Survey, denominational affiliation determines evangelicals through membership in Protestant denominations that are designated as evangelical (which is often used as a synonym for fundamentalist or traditionalist and is distinguished from the moderate and liberal categories of denominations).
This approach depends, of course, on which denominations you label evangelical, and here significant disagreements arise. Some methods include historically African-American Protestant denominations; others do not. So while the former estimate that about 29 percent of the population is evangelical, the latter put the number at around 25 percent.
The Gallup Organization and the Pew Research Center use self-classification as their approach, asking respondents if they call themselves evangelical or born-again Christians. The big uncertainty about this technique is whether born-again and evangelical should be considered roughly synonymous, since the former is largely a matter of experience, and the latter is largely a matter of specific beliefs (such as having a responsibility to share one's faith with non-Christians).
The self-classification approach opens the evangelical category to Catholics and Orthodox Christians as well as to Protestants. Gallup's estimates of the portion of the adult population that is evangelical have ranged from 35 percent in 1976 to a whopping 47 percent in 1998; last year's estimate was 41 percent.
The most restrictive approach, used most notably by the Barna Research Group, takes specific religious beliefs as the determinant of religious identity. Barna, like Gallup, allows non-Protestants to be counted as evangelicals, but the beliefs are so precisely delineated that only 2 percent of Catholics make it into this category (compared with 16 percent of Catholics in the Gallup Poll).
Overall, Barna finds that only 7 percent of the adult population is evangelical.
Knowing what goes into the making of these different polls will not allow you to predict whether Obama will, or will not, be able to make sufficient inroads into the evangelical voting bloc in the coming election. But it is reasonable to expect that he will win more votes from those evangelicals identified as such by Pew and Gallup than those identified as such by Barna.
Why? Because evangelicals defined by specific beliefs tend to be more conservative than those defined by self-classification.
Drawing on data from the 1998 General Social Survey, which had questions using the three ways of defining evangelicals, Lindsay and Hackett found that 53 percent of those determined by self-classification were Republican compared with 63 percent of those determined by specific beliefs. Among other things, that might explain why a recent Barna poll found Obama leading in 18 of 19 American faith communities, the only exception being the very restrictively defined evangelical community.