The Devil in the Details

Polls reinforce reporters' stereotypes about evangelicals

My New York friends congratulated me for my “bravery” when I headed off to cover evangelical supporters of Mike Huckabee’s presidential campaign in Iowa shortly before Christmas. I grew up in Virginia at a time when the state’s Christian right was gaining strength, but have spent most of my adult life in liberal circles where evangelicals—if there were any—kept their faith largely to themselves.

During the days I spent following Huckabee’s campaign swing, I met some Christians who conformed to my friends’ expectations: home-schoolers, knee-jerk fundamentalists, voters for whom “family values” trumped all other issues. But I seemed to meet just as many evangelical voters who defied the stereotype. In Ames, there was the engineering grad student who was concerned about energy independence and was choosing between Huckabee and the libertarian Ron Paul. In Des Moines, a sixty-something woman whispered to me behind her husband’s back that she was uncomfortable with the hard-right line on abortion and gay unions. In Waterloo, I spoke by phone with a member of Huckabee’s “Pastors Council” who mentioned he was African American. At stop after stop, Christian voters—along with a good number of Huckabee supporters who said they were not regular churchgoers—cited economic concerns as often as social ones. And when I stepped off the trail to cover Christmas services at a Charismatic church, I even met evangelical Democrats.

So I was surprised on December 21, 2007, when The New York Times asserted that, “People who have been coming to Mr. Huckabee’s rallies are mainly Christian activists.” I talked to a lot of churchgoers, sure, but they didn’t strike me as any more “activist” than any Iowan who participates in the state’s caucuses. On election night, January 3, I hoped that the exit polls would reflect the diversity I had seen among Iowa evangelicals. But reporters relying on the National Election Pool, the exit-polling consortium of major news outlets, reported that evangelicals or born-again Christians (terms the NEP considers interchangeable) accounted for 60 percent of GOP caucus-goers. Period. There was no mention of evangelical Democrats, no attempt to parse those terms. And the fact that Huckabee defeated his closest rival, Mitt Romney, by twenty-seven points among these evangelical and born-again voters reinforced the perception that his campaign was fueled by passionate “Christian activists.”

Had I totally missed the forest for the trees, overemphasizing the exceptions that proved the rule about evangelical Christians? Maybe. But when I looked closely at the polling myself—both the exit polls and general polling about religion and politics—I found it to be incomplete and confusing, confirming on its surface the assumptions that many reporters carry with them when they parachute into places like Iowa. For instance, reporters would have had a hard time finding evidence of evangelical influence in the state’s Democratic caucuses even if they had thought to look for it—the NEP didn’t ask Democrats about their religious affiliation. This is an especially egregious example of the many ways polling data can help reporters miss the trees from their bird’s-eye view of the forest.

Paul Vitello, the Times reporter who concluded that Huckabee’s audiences were mostly “Christian activists,” told me in an interview, “Iowa is surprisingly homogenous to an overwhelming extent….Out there, I saw absolutely no tension [about excluding people of other faiths around Christmas time], and it was that lack of tension that I found interesting.” He said he did not dig too deeply into the pre-election polling, but had read about conservative evangelicals and their support for Huckabee. His interviews confirmed the impression he had before he set out, he said, as did the returns on primary day.

But there was evidence that Iowa evangelicals—and Huckabee’s coalition—were more diverse than the reporting made it appear, which reporters could have seen had they simply taken more time to scrutinize the exit-polling data. Huckabee actually failed to win a majority of evangelicals. He won a plurality, 46 percent, but the majority split its votes: 19 percent went to Mitt Romney, 11 percent to Fred Thompson, and 10 percent each to John McCain and Ron Paul. At least 20 percent of Huckabee’s support came from nonevangelicals. These data could just have easily supported stories about how Huckabee owed his victory to less-affluent voters, a large portion of whom happened to be evangelical. Huckabee beat Romney by eleven points among the 81 percent of voters with a family income of less than $100,000, and the poorer the voting bloc, the bigger his margin—among those with a family income of less than $30,000, Huckabee beat Romney by thirty-three points.

A feedback loop took hold in the coverage: reporters expected evangelicals to support Huckabee, and there were no prominent red flags in the polling that would have made them question that expectation. Two sets of numbers did come along a month later that caused at least some reporters to put some pressure on this notion of a monolithic horde of conservative Christian voters. One was conducted by Zogby International for the advocacy group Faith in Public Life, along with the Center for American Progress, a liberal think tank. Unlike NEP’s exit polling that failed to ask Democrats whether they were evangelical, FPL conducted exit polls in Tennessee and Missouri, and found that 30 to 40 percent of evangelicals in those primaries voted Democratic. It also found that evangelical voters support—by a margin of more than fifty points—an agenda that includes ending poverty, tackling the aids epidemic, and protecting the environment over one limited to concerns about abortion and gay marriage. This received minor attention from major outlets—CBS and The Washington Post mentioned it on their campaign blogs, for example—but it was largely ignored by the mainstream press.

A handful of smaller news outlets picked up on a more idiosyncratic but equally intriguing set of data from the Barna Group, a Christian market-research firm. “[M]any born-again voters are trading places this election year,” wrote The San Diego Union-Tribune in a February 16 news article, while the Nashville Tennessean headlined its February 9 religion brief GOP losing support of born-agains. Barna found that 40 percent of the born-again voters it interviewed were inclined to vote for a Democratic presidential candidate, while only 29 percent preferred a Republican. (This evidence of a growing “religious left,” as it was portrayed in press accounts, coincided with the release of the latest book by Jim Wallis, a longtime progressive evangelical activist.) But just five-hundred words into Barna’s study, titled “Born Again Voters No Longer Favor Republican Candidates,” was the subhead, “Evangelicals Remain Strongly Conservative Republican,” which these outlets seemed to have missed—or ignored.
Here again, the complexities of the data went largely unexamined, and the complexities tell us as much—if not more—than the broad conclusions. It probably never occurred to most political reporters to wonder about the definition of an “evangelical” or “born-again” voter, especially since NEP and most secular polling organizations do not differentiate. But the Barna Group does. There is no consensus within Christian circles about the meaning of the terms. The National Association of Evangelicals has a statement of beliefs that might define evangelicalism, but it represents denominations that account for only half of the most widely accepted estimates of American evangelicals.

Barna also breaks with most secular polling organizations by classifying respondents based on questions about their beliefs instead of asking them to identify themselves. Barna’s “born-agains” are those who say they have “made a personal commitment to Jesus Christ that is still important in their life today” and believe that “when they die they will go to Heaven because they had confessed their sins and accepted Jesus Christ as their personal savior.” Barna counts them as 42 percent of the population, and this is the group that is leaning Democratic. “Evangelicals” are a small subset of “born-agains” who affirm a set of beliefs derived from the NAE’s credo. They include “believing they have a personal responsibility to share their religious beliefs about Christ with non-Christians; believing that Satan exists; [and] believing that eternal salvation is possible only through grace, not works.” Barna’s “evangelicals” account for 8 percent of the population, and remain solidly Republican.

The organization’s founder, George Barna, is engaged in an ideological battle to enforce a strict evangelical theology. He uses polling to identify those he considers “true” evangelicals. But, despite his agenda, his polling is regarded by secular pollsters as statistically sound, and it has something to teach political reporters: that the religious voters who usually show up as a unified bloc are in fact diverse theologically and politically. Barna’s breakdown shows that different subgroups are reacting in different ways to a shifting political landscape—and religion is just one factor in their decisions. Evangelicals carry their faith in different ways, and are many things besides religious; they are also midwestern, or white, or middle class, etc.

A starting point for reporters covering this story, then, is to be more sensitive to the ways implicit judgments shape the picture of evangelicals that emerges from mainstream polling. The Pew Forum for Religion in Public Life is widely regarded as the industry standard for religious polling. But, for example, when it joins with its sister organization, the Pew Research Center for the People and the Press, to survey evangelical politics, it generally reports only on “white evangelical Protestants.” There are defensible statistical reasons for excluding the 19 percent of evangelicals who are people of color. When Pew has done detailed polling to track political variations within different evangelical demographics, it has found that non-white evangelicals consistently differ from their white counterparts. Also, not enough members of these groups show up in a random sample of 1,500 voters to derive reliable data on their views. But these voters skew much more liberal than white evangelicals and are easily forgotten since they are seldom reported on independently—one in five evangelicals essentially disappears when the faith community’s politics are reported on.

Jeff Sharlet, the editor of The Revealer, an online magazine about religion, suggests that this oversimplified polling allows political reporters to ignore the complicated factors that go into individual voting decisions in order to “get a voting bloc that doesn’t quite exist.” Historically, there is nothing inherently Republican about evangelicals, even white evangelicals. The Democrats and Republicans could claim roughly equal evangelical support in the late 1980s, but it skewed heavily GOP during the Clinton and Bush years, a trend that now shows signs of reversing. This was due to several factors, including the more complete capture of white southerners (who were heavily evangelical) by the Republican Party leading up to the 1994 election, the intensification of the “culture wars” under Bill Clinton, and George W. Bush’s religious rhetoric and emphasis on faith initiatives. But this percolating realignment is complicated, too. So the rise of a “religious left,” Sharlet says, “is not there. It’s only coming to people who didn’t understand the religious right.”

The secular press started covering evangelical Christians when the spiritual movement’s impact began to show up at the ballot box in the 1980s. This caused many to inaccurately equate conservatism with evangelicalism, a diverse faith community to which one in four Americans belongs, according to most estimates. The New York Times went so far as to fold coverage of evangelical politics into a “conservative beat” for the 2004 campaign. As Sarah Posner noted last fall in her FundamentaList blog for The American Prospect, Bush’s victory in 2004 exacerbated the problem, leading the mystified secular press to portray “‘values voters’ (read: biblical conservatives) as the homogenous soul of American evangelicalism.” The growing strength and visibility of evangelical moderates—and evangelical liberals—is an important story in the 2008 election. But these voices have long been present, even if they have mostly fallen on deaf ears in the mainstream press. We may get this new twist in the ongoing story wrong, too, if we don’t look more closely at the data, especially when they seem to confirm what we think we already know. 

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Lester Feder is a freelance reporter based in Washington, D.C., and a research scientist at George Washington University School of Public Health.