As a candidate, Donald Trump’s views on Syrian refugees left little to the imagination. In the final debate, he called Syrian refugees “definitely, in many cases, ISIS-aligned...the great Trojan horse.”Footnote 1 Religious leaders in the United States struck a different tone. Whereas religious groups vigorously debate one another on a great many issues (e.g., abortion, capital punishment, climate change, immigration, Israel/Palestine, and rules for bathrooms and wedding cakes), on the Syrian refugee crisis, leaders across religious traditions invoked sacred texts to stake out similar positions. Statements from Catholic, Presbyterian, Pentecostal, Methodist, Episcopalian, Lutheran, Mormon, Jewish, and Muslim groups all provided explicitly religious arguments for supporting refugees. Many specifically argued for bringing more Syrian refugees into the United States (see the online appendix for details).
Even evangelical groups and their leaders—often among the most politically conservative voices in American politics—joined the chorus, potentially cross-pressuring evangelical Republicans. The National Association of Evangelicals asked Congress to “expand expedited resettlement in the U.S. of vulnerable Syrian refugees.”Footnote 2 The Southern Baptist Convention, the largest evangelical denomination and a conservative bulwark, issued a resolution in 2016 that supported refugee resettlement in the United States.Footnote 3 To be sure, some evangelical leaders opposed resettling Syrian refugees in the United States. Most notably, Franklin Graham—CEO of the evangelical relief organization Samaritan’s Purse and the son of the revered evangelist, Billy Graham—supported Trump’s executive order that temporarily suspended the refugee-resettlement program. Still, evangelical leaders overwhelmingly broke with Graham: more than 500 evangelical pastors and leaders signed an open letter to “call on President Trump and Vice President Pence to support refugees” published in the Washington Post (Weber Reference Weber2017). As Green (Reference Green2017) stated, “[f]rom religious leaders’ perspectives, backlash against Trump’s immigration policy may be the most ecumenical issue in America right now.”
For many issues, partisan and religious forces align. For example, evangelicals can encounter conservative religious cues on a host of issues (e.g., abortion, religious liberty, and LGBT rights) that match the conservative messages from Republican sources. However, on the question of letting Syrian refugees into the United States, religious elites’ generally pro-refugee messages ran counter to many conservative and Republican voices, especially Trump’s. In this case, what is the relative importance of explicitly political forces like party identification and ideology vis-à-vis religious forces? Exploring this question provides a first step toward understanding who supports bringing Syrian refugees to the United States. We know relatively little about the underpinnings of attitudes toward Syrian refugees in the United States. Given the scale of the refugee crisis, its impact on global politics, and the political significance of Trump’s executive orders limiting refugee resettlement in the first months of his administration, refugee politics deserve scholarly attention.
However, on the question of letting Syrian refugees into the United States, religious elites’ generally pro-refugee messages ran counter to many conservative and Republican voices, especially Trump’s.
RELIGIOUS AND PARTISAN FORCES
Religion directly and indirectly shapes public opinion on a variety of social, economic, and foreign-policy issues (Guth Reference Guth2013; Jelen Reference Jelen, Guth, Kellstedt and Smidt2009; Wilson Reference Wilson, Guth, Kellstedt and Smidt2009) often by highlighting values, information, and cues relating to those issues. Values, information, and cues can be transmitted and reinforced via statements from clergy or other leaders during services or in religious media outlets. However, although clergy can shape the laity in some ways, direct influence often is limited (Djupe and Calfano Reference Djupe and Calfano2013; Djupe and Gilbert Reference Djupe and Gilbert2009). Cues from interactions with co-religionists at worship services and in less formal gatherings associated with the local house of worship often can be more influential (Djupe and Calfano Reference Djupe and Calfano2013; Djupe and Gilbert Reference Djupe and Gilbert2009).
Of course, political forces such as party identification, ideology, and affect toward political leaders powerfully influence views on various issues as well (Abramowitz Reference Abramowitz2010). Thus, I expected those identifying as Republican, conservative, and/or Trump supporters to be less likely to approve of bringing refugees into the United States. Given the almost universally pro-refugee messages from religious leaders across traditions, I expected few differences in attitudes toward refugees across religious traditions.Footnote 4 In addition, I expected that, all else being equal, people who regularly attend worship services would be more favorable toward allowing refugees into the country. Regular attenders presumably are the most likely to encounter pro-refugee messages that they may hear during services. Attendance also may proxy engagement with their tradition’s leaders and members outside of services (e.g., via religious media and informal meetings with co-religionists). Regular attenders also may hold more strongly to values that encourage sympathy for refugees—the same values that may be animating their leaders’ pro-refugee statements. Given that leaders from various traditions voiced pro-refugee messages, I expected regular attenders to be more favorable toward refugees regardless of religious tradition.
Despite broad consensus among religious leaders, mass opinion broke down along religious lines. Table 1 shows support for allowing Syrian refugees into the United States across three surveys. A Pew Research Center poll from September 2015, soon after President Obama committed to allowing 10,000 Syrian refugees into the country, asked respondents whether they approved of allowing more Syrian refugees (see the online appendix for details on the measures). The table shows the percentage that approved, using Pew’s religious tradition measures. The survey revealed significant religious differences, with approval ranging from 70% of Jewish respondents to 31% of white evangelical Protestants. Despite religious leaders’ calls for supporting Syrian refugees, those unaffiliated with any religion were more supportive than those among many of the major US religious traditions.
Table 1 Support for Allowing Refugees in the United States by Religious Tradition
Table 1 also presents findings from two American National Election Studies (ANES) surveys: the pilot study conducted in January 2016 and the time-series pre-election study fielded from September through the day before the election. The surveys asked how much respondents favored or opposed (or neither) “allowing Syrian refugees to come to the United States,” creating a 7-point scale with strongest opposition at the scale’s low end. Means for religious traditions (constructed to match the Pew study as closely as possible) varied considerably. White evangelical Protestants were the least supportive of allowing Syrian refugees into the country, with a mean about 0.75 points lower than the overall mean, whereas the religiously unaffiliated were among the most supportive, with a mean about 0.50 points higher than average.
These differences mostly disappear when controlling for political variables. Using the Pew survey, I estimated a probit model of approval for allowing more Syrian refugees into the United States as a function of religious (i.e., religious tradition and attendance at a local house of worship), political (i.e., party identification, ideology, and support for Trump), and demographic (i.e., college education, age, income, and race/ethnicity) variables. Religious tradition, education, race/ethnicity, and age variables are indicator variables that are mutually exclusive and exhaustive within the relevant category. Party identification and ideology were coded such that higher scores were Republicans and conservatives. Support for Trump equaled 1 if respondents listed Trump as their first or second choice for the Republican nomination and 0 otherwise.
The ANES models take the 7-point scale described previously as the dependent variable and use the same independent variables with the exception of support for Trump. These models include respondents’ feeling-thermometer rating of Trump rather than their preference for the Republican nomination. I estimated the model using OLS for ease of interpretation but ordered probit models generate similar results.Footnote 5 I recoded all independent variables to a 0–1 scale.
The results supported expectations. In all three surveys, Republicans, conservatives, and Trump supporters were significantly less supportive of resettlement. Consistent with other surveys, those with college degrees and younger people were more supportive.Footnote 6 Compared to a pure independent, moderate, with mean support for Trump in the ANES pre-election survey, the model estimates a strong Republican, strong conservative giving Trump an 85 rating (90th percentile) would be 2.4 points lower on the 7-point scale.
As expected, holding constant party identification, ideology, Trump support, education, and age, relatively few differences remain across religious traditions. In the Pew survey, Mormons and the religiously unaffiliated were significantly more likely to approve than others. In the ANES pre-election survey, Mormons, Catholics who were not white or Hispanic, and respondents affiliated with smaller religious traditions in the United States were significantly more favorable toward refugees than average. (The same was true for adherents of smaller religions in the pilot study.) No other religious tradition’s parameter estimate was statistically significant at even the 0.10 level. That is, conservative evangelical Republicans appear to be no different from other conservative Republicans on this issue. Conservative Republican white Catholics likewise appear indistinct from other conservative Republicans, and so on. The Mormon community stands out from other traditions in two of the three surveys, a point I return to below.
Yet, religion still plays the anticipated role, as those who regularly attend religious services tend to be more supportive of refugees. However, the magnitude of religion’s predictive impact is much smaller than that of the political variables. Compared to non-attenders, those who attend more than once a week were 11 percentage points more likely to support allowing Syrian refugees into the United States in the Pew survey and about 0.45 points more supportive on the ANES 7-point scale. This 0.45-point difference pales in comparison to the 2.4-point difference associated with partisan differences noted previously. I examined whether the attendance estimate differed across religious traditions via interactions between attendance and religious tradition and found that it did not (see the online appendix).
Compared to non-attenders, those who attend more than once a week were 11 percentage points more likely to support allowing Syrian refugees into the United States in the Pew survey and about 0.45 points more supportive on the ANES 7-point scale.
PARTISAN NEWS
The ANES pre-election survey provided a closer look at two sources of partisan cues: Fox News and MSNBC. According to data from the Global Database of Events, Language, and Tone (GDELT) Television Explorer database, from September 1, 2015, to November 8, 2016, Fox News mentioned “refugee” or “refugees” 16 times a day, on average, whereas MSNBC mentioned the terms 12 times a day.Footnote 7 Viewers of either source presumably would have encountered discussion of refugees at various times during the campaign. Although a full content analysis of coverage is beyond the scope of this article, it seems safe to assume that Fox News coverage was generally opposed to allowing Syrian refugees in the United States and that MSNBC coverage was more favorable. Fox News mentioned refugees within four sentences of the words “terrorism,” “terrorist,” or “threat” 1,352 times during the same period, compared to 874 times for MSNBC. In contrast, refugees were mentioned in proximity to “humanitarian,” “victim,” or “victims” 1,375 times on MSNBC and only 979 times on Fox News.
ANES pre-election respondents who said they heard about the presidential campaign on television were given a list and asked to indicate which television shows they watch at least once a month. I created an indicator variable for respondents who watched Hannity, The O’Reilly Factor, or The Kelly File, each of which aired on Fox News. I created another indicator variable for respondents who watched All in with Chris Hayes, Hardball with Chris Matthews, or The Rachel Maddow Show, each on MSNBC.
Remarkably, even after controlling for party identification, ideology, support for Trump, religious attendance, age, and education, consumption of Fox News and MSNBC differentiated respondents’ views of refugee resettlement (table 2, column 4). Ceterus paribus, Fox News viewers were about a quarter-point less favorable toward allowing refugees into the United States and MSNBC viewers were about a quarter-point more favorable. Although these cross-sectional data do not support causal inferences, the results are consistent with a causal impact because the model controls for so many factors that might affect media choice and other studies have demonstrated partisan media’s causal impact (e.g., Levendusky Reference Levendusky2013).
Table 2 Support for Allowing Refugees in the United States
Notes: Standard errors in parentheses. ***p<0.01, **p<0.05, *p<0.1. Marginal Effect refers to the average change in probability of approving when varying the value of the given variable from minimum to maximum (0 to 1), as calculated by STATA’s margins routine.
Comparing means on the 1–7 scale among white Democratic identifiers across the largest religious traditions shows the difference between MSNBC viewers and non-viewers. Although limiting the analysis to white Democrats and then parsing by MSNBC viewership and religious tradition is slicing the data fairly thin, remarkable differences are evident. Figure 1, which shows mean responses along with 95% confidence intervals, shows that MSNBC viewers were distinctively favorable toward Syrian refugees, controlling for race and party identification. Among white evangelical Democrats, MSNBC viewers were about one point more favorable toward refugees than non-viewers; however, there were so few respondents in these categories that confidence intervals overlapped considerably. Religiously unaffiliated white Democrats who watched MSNBC were 0.8 points more supportive than non-viewers of resettling Syrian refugees in the United States, although confidence intervals just overlap. More starkly, MSNBC viewers among white mainline Democrats were 1.7 points more favorable toward Syrian refugees than non-viewers, whereas white Catholic Democrats who watched MSNBC were slightly more than one point more favorable than non-viewers in this group.
Figure 1 MSNBC and Views toward Refugee Resettlement among White Democrats
We see the reverse pattern for Fox News viewers and white Republicans (figure 2). In particular, among white evangelical Republicans, Fox News viewers had an average of 1.7—a full 1.5 points below the overall mean and 0.8 points below their fellow white Republican evangelicals who did not watch Fox News. If church attendance and watching Fox News are competing sources of information and cues, then Fox News is winning.
Figure 2 Fox News and Views toward Refugee Resettlement among White Republicans
Other differences shown in figure 2 are less dramatic but still significant. Among white Republican mainline Protestants and Catholics, Fox News viewers scored about 0.6 points lower than their non-watching counterparts. For religiously unaffiliated white Republicans, the difference was a statistically insignificant 0.4 points. In summary, even among white Republicans, Fox News viewers stand out for their opposition to allowing Syrian refugees into the United States.
Democrats, liberals, Trump skeptics, MSNBC viewers, college graduates, young people, Mormons, and regular church attenders were all more favorable than their counterparts toward bringing Syrian refugees into the United States. That said, however, party identification and ideology were more predictive of attitudes toward refugees than attending religious services.
CONCLUSION
Democrats, liberals, Trump skeptics, MSNBC viewers, college graduates, young people, Mormons, and regular church attenders were all more favorable than their counterparts toward bringing Syrian refugees into the United States. That said, however, party identification and ideology were more predictive of attitudes toward refugees than attending religious services. These analyses take an initial step toward understanding refugee politics in the United States and raise important questions. First, exactly which messages about refugees affect public opinion? To what extent is opposition to resettlement driven by anti-Muslim sentiment, fear of terrorism, economic worries, and/or concern about weakening American culture? I assumed much about the content of, exposure to, and acceptance of religious and political messages—assumptions that are, I think, reasonable and probably correct but almost certainly overgeneralized. Although suggestive, these analyses cannot establish causal links. Future research should provide more precise measures of message content and use experimental methods to gauge reactions to those messages.
Why might partisan and ideological effects outweigh religious forces on attitudes toward refugee resettlement? A simple reason may be that partisan messages are so clear, strong, and frequent—compared to religious messages—that for some (perhaps many) Republicans, religious and partisan cues may not have been in much conflict. Individuals may be wholly unaware of religious leaders’ statements, which may not be widely reported in news media or at the local level. Moreover, at least among Protestants, few pastors discussed the issue at all during worship services. In a 2016 survey of Protestant pastors, only 32% of evangelical pastors and 41% of mainline pastors said they had ever “specifically addressed the Syrian refugee crisis from the pulpit” (LifeWay Research 2016). Furthermore, local pastors may not share their leaders’ views on refugee issues. In the same survey, 63% of Baptist pastors said the United States “can balance national security and compassion in its response to global refugees,” meaning a significant minority may have sided more with Franklin Graham than the Southern Baptist Convention leadership. Even if people know about their leaders’ pro-refugee statements, they presumably are not being reminded of them frequently. In contrast, Fox News discussed refugees and the refugee crisis several times a week during the 2016 campaign. Finally, Fox News may have minimized the conflict between religious and partisan cues by highlighting Franklin Graham’s opposition to resettling refugees. Fox News mentioned the evangelical leader at least once on 38 separate days during the period of study, compared to only four days on MSNBC.
As mentioned previously, the Mormon community was distinctively supportive of refugees. Mormons’ uniqueness is consistent with many of the suggestions in the previous paragraph. Compared to other religious traditions, the Mormon tradition has a clearer leadership hierarchy with clear channels of communication and members of the Mormon community tend to pay closer attention to leadership messages and shape their attitudes accordingly (Campbell, Green, and Monson Reference Campbell, Green and Quin Monson2014).
Future research should continue to explore attitudes toward refugees and the conditions under which religious forces could unify or further polarize the public on these and other issues. On many issues, religious cues reinforce partisan and ideological cues, potentially polarizing the public. The similar responses of many religious leaders to the Syrian refugee crisis could bridge a major partisan divide. So far, they have not.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/S1049096518000562
ACKNOWLEDGMENTS
The author thanks Chris Soper for his suggestions.
