"Religious identity serves as a language of social and political interpretation," said Thomas A. Hirschl, Cornell professor of development sociology and first author of the study being published in the December issue of Social Sciences Quarterly (90:4). "But this interpretation depends upon the material circumstance of the interpreter."

The researchers -- who also include James Booth, Cornell professor of biological statistics and computational biology, and Leland Glenna, a sociologist at Pennsylvania State University (who has a master's degree from Harvard Divinity School) -- found, for example, that "white support for Republicans is fractured by religious tradition, biblical authority, social class and gender," while black support for Democrats is equally strong across religious tradition, biblical authority, social class and gender, and "in no way resembles the fractured pattern of white religious partisanship."

In looking at social class, "wealthy interpreters see things differently than poor interpreters," said Hirschl.

The tendency of Protestants, Catholics and others to vote Republican, increases with income, the researchers found. Specifically, wealthy biblical literalist Protestants vote strongly Republican, while low-income biblical literalist Protestants still favor Republicans, but much more weakly.

On the other hand, biblical literalist, low-income Catholics tend to vote Democratic; and the link between wealthy Catholics and the Democratic Party is much weaker.

"We believe this reflects differing interpretations of the Bible, where lower-class Catholics, versus Protestants and upper-class Catholics, believe the Bible advocates social redistribution, and that this interpretation is associated with a preference for the Democrats," said Hirschl.

As for gender and religious identity's influence on voter's choice, men favor the Republican Party more than women. "Women tend to want more government programs to support children, child care and social services, but this varies by biblical belief, race and social class," write the authors.

"Religious identity matters in politics," conclude the researchers, "but its effect is contingent on the individual's social location."

Hirschl added that resolving such social questions as poverty and childhood hunger "may require speaking across religious and secular worldviews, and then agreeing on a course of action. Our research suggests that this should be possible given that religious, as well as secular, interpretations of society are strongly influenced by economic and social circumstances."

The study is based on the authors' analysis of data on 5,543 individuals from the General Social Survey that covered six presidential elections, from 1980 to 2000.

Provided by Cornell University (news : web)