Religion is sometimes used as shorthand to define a political candidate's views -- "evangelical" and "antiabortion" may be considered interchangeable. But a new analysis by political scientists at the University of Houston suggests party labels are far more powerful predictors of how voters perceive candidates.
"Party brand matters," said Elizabeth Simas, assistant professor of political science. "You're running under a label."
Simas and doctoral student Adam Ozer say their findings, published in the journal Research & Politics, suggest the conventional wisdom and scholarship overstate the role religious affiliation plays in voter perceptions.
It's not that a candidate's religious affiliation doesn't matter to voters, who may use religion as a proxy for trustworthiness or likeability. But when it comes to voter beliefs about where a candidate stands on the issues, as well as how liberal or conservative voters perceive a candidate to be, the researchers found party affiliation is the overriding factor.
"Though Evangelical and Catholic cues do impact impressions of a candidate's stance on abortion, the partisan cue dominates perceptions of overall ideology," they wrote. "These findings further demonstrate the power of the party brand."
The relationship between "cues" -- party affiliation, religious affiliation, race and other traits -- is complex, Ozer said. Previous research had found religious labels affect how people perceive a candidate's stance on issues, but Simas and Ozer say those studies often considered religious affiliation in a vacuum, rather than coupled with party affiliation. Once party affiliation was added to the equation, it became the dominant factor.
"Partisanship is a hell of a drug," Ozer said.
There are practical applications to the finding: "If you are a Democrat in Texas and want to appear more conservative, saying you go to church won't necessarily make people think you are more conservative than any other Democrat," Simas said.
She and Ozer based their conclusions on a review of previous literature, coupled with an analysis of data drawn from a nationally representative sample of 1,008 individuals.
"Our results question earlier assertions about the inferences drawn from religious labels," they wrote. "We offer more evidence for the argument that the religion of a candidate is important, not because it serves as an ideological dog-whistle, but because it is a signal of non-policy characteristics."
Religion, it turns out, is one piece of information among many considered by voters. "Religion is a cue, but it's not the cue it's considered to be," Simas said. "It doesn't tell us much."