For a generation of political scientists, religion has been the great white whale of American politics: They know it’s there, but they’re not sure exactly how to capture it. In particular, they’ve had a special problem determining what role religion plays in shaping the behavior of lawmakers.

One barrier has been the lawmakers themselves. A lawmaker’s incentive to declare a religious affiliation in this devout nation makes such personal attestations suspect, at least in the minds of many academics. Yet to understand the role that faith plays in Congressional voting behavior, researchers first need a clear picture of the faith of the Member who votes.

Now, a group of political scientists has come forward with what they believe to be the first true snapshot of religious faith and practice among lawmakers. Their paper, “The Confessional Congress: Religion and Legislative Behavior,” studies religious faith and voting behavior among House Members in the 105th Congress in 1997-1998.

Depending on one’s prior assumptions about the influence of religion on the political process, the results will either undercut conventional wisdom or confirm the obvious. What the scholars found is that even after factoring out partisan ties and geographical imperatives, religious belief — independently — helps to predict how a Member will cast his or her vote.

In other words, when neither party label nor constituency (nor any number of other variables) did not explain votes taken by a Member, religious belief did.

“Much of religion’s influence is obviously channeled indirectly through changes in the religious composition of the two parties, changes which closely parallel those transforming the nation’s mass electorate,” the authors of the study write.

“Nevertheless,” they add, “even when party membership, district religious characteristics, and important political variables are controlled, religious affiliation and theological perspective still have a direct impact on members’ voting.”

This is more significant than it may sound at first, because it could shake up the prevailing academic orthodoxy about what drives voting by lawmakers. And it seems to fit well with the tectonic shifts driving the ongoing realignment of American politics.

“What we were surprised with, to some extent, is that we anticipated that religion would have more of an influence in what we call a ‘composition effect,’” said James Guth of Furman University, a co-author of the study.

What that means in layman’s terms is that once they’d determined the religious faith of individual Members, the researchers expected to begin seeing associations between religious faith and political party.

This turned out to be true for Members of similar religious participation — more “observant” Members trended GOP, for instance, mirroring the wider population. But after that, no clear patterns of “compositional effect” emerged.

For the study, the authors took the information they collected about the religious beliefs of Members and combined them with other variables, such as age or race, in multiple regression analyses — a statistical method which tests a cluster of possible variables together to determine which of them can explain the behavior observed.

To distinguish “conservative” voters from “liberal” ones, the resulting tables showed the likelihood that the input of a particular variable could predict an identifiable “conservative” vote in a number of areas, such as social issues, economic issues and foreign policy.

The strongest relationship the authors discovered was between evangelical Protestantism and conservative voting. They found a firm connection across the full range of categories: There was a relatively high degree of likelihood that an evangelical Protestant, even if he or she was a Democrat, would vote in favor of the conservative position.

Conversely, the authors found that black Protestant Members were, across-the-board, the least likely to cast an ideologically conservative vote. That is to say, there was a greater likelihood that a black Protestant lawmaker would vote against the conservative position than there was for Hispanic Catholics or Jewish Members to do so, though not by much.

These results might seem to make sense on their face. But consider the Church of Jesus Christ of Latter-Day Saints, known more commonly as the Mormons.

Practicing Mormons do not smoke or drink, and they are regarded as one of the country’s most politically conservative religious groups. Nearly every Mormon sent to Congress is a Republican. Yet Guth’s study showed that Mainline Protestants, who are concentrated in the Northeast and are divided fairly evenly between the parties, were more likely to take the conservative position in every area except foreign policy.

The results surprised some scholars who expected that party membership would more or less explain the way Members voted.

“I personally thought, ‘Well, all this is going to be washed away when they control for party,’” said Geoffrey Layman, a political scientist at the University of Maryland who has been studying the report.

Layman said he expects scholarly “quibbles” with some of the methodology used in the study, which has yet to be published.

He cited, for instance, the authors’ attempt to discern the “theological orientation” of Members — that is, the degree to which “traditional,” as opposed to “modern,” styles of worship are practiced. (Guth agrees this is a likely area of debate.)

But Layman said that if the findings withstand challenges from other scholars, then the conclusions are significant.

“It strikes me that one thing it tells us about Members of Congress is that they do tend to vote their consciences,” he said.

Layman said the findings may also signal a shift from an “affiliation model” of party membership to a “culture wars model.”

Catholics, for instance, were once predominantly members of the Democratic Party coalition. The study suggests that their allegiance is now more or less defined by the way in which they observe the faith, with more traditional Catholics moving to the GOP.

‘Tricky’ Empiricism

Guth and his main co-author, Wheaton College political scientist Lyman “Bud” Kellstedt, presented their findings to the Midwest Political Science Association in early April.

Guth, who is considered a leading expert on Southern Baptists in the political sphere, has been studying the role of religion in political life for roughly 15 years, often with a group of three other scholars that includes Kellstedt, John Green and Corwin Smidt.

This group is renowned for work it has done to study the political universe of evangelical Christians. They have also documented political phenomena such as the decline in Mainline Protestants in Congress and the commensurate rise in smaller denominations.

“They’re really highly respected political scientists,” said Georgetown University political scientist Clyde Wilcox, a specialist on the intersection of religion and politics.

While there has been a vast amount of scholarship about the role of religion in American politics, political scientists have produced little useful information about the influence of religious belief on the behavior of legislators.

Previous academic studies lacked controls and precision. Even conducting in-depth interviews about religious faith with scores of Members of Congress — a tactic used by two researchers in 1982 — did not necessarily improve the quality of the results, scholars say, since no controls could be imposed and the experiments could not be reproduced.

Wilcox, who has not yet seen Guth’s study, said scholars have been limited by a “tricky empirical question.” “Sorting out, for example, how much a Catholic votes their own religion versus the religion of their district has some very, very funky math to it,” Wilcox said.

Rep. Ron Lewis (R-Ky.), an ordained Baptist minister, said that while he would advise members of his church against drinking and smoking, his district produces both tobacco and Jim Beam and Maker’s Mark bourbon. Lewis’ responsibility, as he sees it, is to protect the livelihoods of his constituents, regardless of his personal beliefs.

“I represent 675,000 people. Some are Baptists, some are not. Some are Christians, some are not,” he said. “I have to represent all of them.”

It was with his liquor-industry constituents in mind that Lewis sought not long ago, during a budget surplus, to eliminate the tax on distilled spirits. The effort did not go over well with many fellow Baptists in his district, who viewed the levy as, quite literally, a “sin tax.”

On the other hand, Rep. James Clyburn (D-S.C.), who spearheads efforts by the House Democrats to enhance the Caucus’ relationship with the religious community, described his own political outlook as “biblically based.” He said, for instance, that his views on Social Security policy are based on religious teachings about obligations to widows and orphans.

Members of the African Methodist Episcopal church, a predominantly black denomination, “don’t hesitate to introduce religion to politics,” Clyburn said. He pointed out that the AME church, to which he belongs, was founded as a protest against the segregationist practices of the post-Civil War Methodist Church.

“I’d have to think about whether [religion is] the first determinant,” Clyburn said. But he added, “It’s always, ‘If I can ground that issue in religious themes, that’s what I do.’”

Google Meets Congress

A Roll Call study of the 109th Congress found at least 27 different Christian denominations represented in the House and Senate. But Guth would suggest that such a number is misleadingly small.

Differences in forms of worship and belief within several of the Mainline Protestant churches divide them into what are essentially smaller religious groupings that are more or less isolated from one another.

So, broad affiliations such as “Congregational” or “Lutheran” or “Presbyterian” are unhelpful, since so much variation exists within the churches themselves. ELCA Lutheran and Lutheran Missouri Synod, for instance, share the same religious antecedent, but the latter is Mainline and the former Evangelical.

So reaching conclusions about the religious views of any one Member — let alone the entire House — is daunting. The local church an observant Member attends could be a significant piece of evidence in making conclusions about about the lawmaker’s theology. Even membership in outside groups — such as a Catholic Member’s association with the Knights of Columbus — could be useful.

“We found, for example, that members often report much more specific religious information on their congressional or campaign Web sites,” the authors write, describing their “unobtrusive” and “data-heavy” approach. “Many name their local house of worship, and some detail of personal religious activity, such as holding leadership positions in their congregations, serving on committees or task forces, or teaching Sunday school.”

Internet databases and search engines such as Google also provided a “treasure trove” of information. Internet search results, say the authors, “ranged from newspaper articles in which reporters discussed religious matters with individual members (in two especially helpful cases, with all the members from Tennessee and Wisconsin), to accounts of private devotionalism, such as small prayer groups on Capitol Hill (National Prayer Breakfasts don’t count), to published income tax returns showing substantial religious contributions.”

These resources, the scholars say, enabled them to make reasonably sound judgments about the way Members of Congress worshiped and what their theological beliefs were. Or whether they worshiped at all.

They found that two-thirds of the Members who could be called “highly active” in practicing their faith were in the GOP. Meanwhile, three-fifths of the lawmakers who the researchers determined to be “just members” of a religious group — or who did not discernably practice a religious faith — were Democrats.

Moreover, the Members who could be classified as “traditionalist” also tended to be the most actively observant.

Yet when Guth and his colleagues plugged the data into their multiple regression tables, the intensity with which Members practiced their faith disappeared as a significant factor in predicting how they would vote.

“This suggests, of course, that religious traditionalism tends to produce both greater activity among its adherents — and conservative legislative voting,” the authors write.

Take the Mainline Protestants. On foreign policy, the authors found that a traditionalist outlook on theology predicted conservative votes across all categories. And a modernist outlook — at the other end of the theological spectrum — worked just as well to predict votes against the conservative position.

Since the authors had controlled for other possible explanations, the results suggested that differing theological perspectives helped to explain a divergence in ideology. “Religion clearly matters in American politics, and we need to know much more about how it matters on Capitol Hill,” Guth and his co-author write.

Joanna Stein contributed to this report.