Religious Rhetoric and Religious Identity for Non-Christians
Religious cues can be a powerful rhetorical tool to induce positive emotions and activate identity-relevant concerns in Christians; however, non-Christian Americans have a markedly different response. To begin, note that not all non-Christian respondents have the same reaction to religious cues. Although non-Christians from a variety of religious affiliations (e.g., Judaism and Islam) tended to have significantly lower levels of support for candidates who used religious language, this drop was particularly pronounced for individuals with no religious affiliation. For individuals who did not identify with any organized religious tradition (primarily self-identified agnostics and atheists), civil religion rhetoric appears to be a particularly substantial turnoff. This nosedive in support among nonidentifiers makes sense. Even though civil religion has been described as a “minimal monotheism,” it is theism nonetheless. Given that civil religion purports to be speaking for all Americans, it makes sense that those who remain excluded from this vision of the American community would evaluate the candidate negatively. Among individuals from non-Christian faith traditions, the subgroup priming actually improved support compared to the control. This suggests that the mere mention of multiple faith traditions in this condition was enough to generate broad-based support, even when the traditions mentioned were not their own.28
Just as previous findings illustrate that Christians exposed to civil religion rhetoric are ultimately primed to evaluate the candidate on different terms, so too did non-Christians come to see civil religion candidates in a fundamentally different light. I examined this by looking at both the traits ascribed to the candidates and the emotions that the rhetoric induced.29 Two findings stand out. First, non-Christians tended to view candidates in both the civil religion and subgroup religion conditions as more religious. But after being primed with a civil religion cue, the perceived candidate religiosity had a strong negative correlation with vote choice (r = -.425, p < .01). Thus, among non-Christians in the sample, a strong civil religion candidate was viewed as being more of a liability than an asset. Perceived religiosity was not, however, a liability in the subgroup religion condition.
This pattern can be explained by the specific identities cued in the experiment. The subgroup religion priming made reference to multiple faith traditions: “No matter what your racial, ethnic, or religious background—whether you’re a Baptist, a Catholic, a Pentecostal, or whatever—these values are enduring.” Even when the non-Christians did not fall into any of these denominational groups, this rhetorical style was apparently inclusive enough such that religiosity was not made germane to vote choice. For non-Christians, subgroup and civil religion rhetoric appear to give different impressions of the candidate’s own religious tolerance. Subgroup religion rhetoric signals a pluralistic and open religiosity, whereas civil religion rhetoric appears to signal a less tolerant religiosity that is threatening to non-Christians. Thus, religious rhetoric always makes candidates appear more religious to audiences, but the specific religious image that a candidate projects makes a substantial difference in whether it will garner support or create opposition among non-Christian constituents.
In addition to seeing subgroup and civil religion candidates as more religious, non-Christians also had a different emotional response to the religious cues and made different inferences about the candidate’s empathy.30 More than 85 percent of non-Christian respondents in the control condition felt that “caring” described the candidate either “quite a bit” or “extremely well.” In contrast, only 67 percent of respondents in the civil religion condition and 60 percent of respondents in the subgroup religion condition felt similarly. Likewise, when candidates invoked the genre of American civil religion, non-Christians reported feeling significantly more irritated and less proud, hopeful, and excited.3 1 Although a candidate compassion gap was also present in the Christian sample, it was considerably smaller. This large Christian/non-Christian difference says volumes about how religiously infused leadership is understood by the mass public. Relatively nondescript campaign platitudes lead non-Christian voters to see candidates as less compassionate, even when the rhetoric is gesturing toward inclusivity. Non-Christian voters react to civil religion and subgroup religion cues in a dramatically different way than Christian voters. Whereas civil religion rhetoric has tremendous appeal among Christians (particularly those who scored high on the CRI scale), it can be a liability among non-Christian Americans, inducing negative emotions and creating the impression that the candidates lack compassion. Religious rhetoric need not speak of a culture war to foster divisions.