______
Faith, Source Credibility, and Trust in Pandemic Information
JIANING LI, AMANDA FRIESEN, AND MICHAEL W. WAGNER
One of the greatest mass communication challenges during the COVID-19 pandemic was information fragmentation. While most news organizations were reporting about the contagiousness of the virus and strategies to prevent becoming infected with the virus, some outlets, like Fox News, were downplaying the severity of COVID-19. Sean Hannity regularly told his viewers that predictions about the seriousness of the virus were “wrong” and expressed doubts about taking a vaccine. Laura Ingraham tweeted that the pandemic was a good time to fly (though she later deleted the tweet). One of the leaders of the federal government’s Coronavirus Taskforce, Dr. Anthony Fauci, called some of Fox’s reporting “outlandish” (Stelter 2020). Of course, it wasn’t just Fox News spreading misinformation about COVID-19. The president of the United States, Donald Trump, was regularly minimizing the severity of the pandemic and promoting untested and potentially dangerous treatments and alleged preventive behaviors. While there were times that the president indicated the virus was serious and that people should wear masks, he usually quickly pivoted back to downplaying the virus’s severity.
There was considerable variation at the state government level as well. For example, the Republican-led state legislature in Wisconsin took the state’s Democratic governor to court, overturning a stay-at-home order. In short, Americans heard mixed messages about COVID-19 from different levels of government, different branches of government, the two major parties, and news sources. Of course, people do not consume news coverage in a vacuum. People’s news use is partially driven by predispositions and individual identity characteristics (Shah et al. 2017). As such, how individual identities are related to the trust people have in leaders, news sources, and scientific organizations that provide information about COVID-19 is important to understand. One identity that has the potential to shape how individuals evaluate their major sources of information about the pandemic is their religious identity. In particular, one’s religion and their religiosity may be related to the trust and attention people place in various sources related to the pandemic.
Moreover, faith leaders exhibited considerable variation in the messages they gave to their flocks about the coronavirus. One Maine pastor, presiding over what National Public Radio (2020) called a “superspreader wedding event,” attacked public health officials for intervening in church business, saying, “You’re looking at a liberty lover! I—love—liberty. And I want the people of God to enjoy liberty” (All Things Considered 2020). In Wisconsin, a coalition of faith leaders called on the state’s legislature to do more to clear up confusion and help keep their parishioners safe (Gunn 2020).
Who did Americans of different faith traditions trust to tell them the truth about COVID-19? In this chapter, we provide evidence about the answers to these questions using an online national survey that asks about respondents’ religious faith, media use, media trust, and trust in various leaders to tell the truth about COVID-19. We show that even though there were not meaningful differences about how much people trusted their most trusted news source to tell them the truth about the coronavirus, evangelical Christians are more likely than others to rely on news sources that provide mixed messages about best practices during the pandemic. We also find that identifying as an evangelical Christian and holding literal views of the Bible were positively associated with trusting Donald Trump to provide good information about the pandemic. Even so, levels of trust were higher for the Centers for Disease Control than the president with respect to COVID-19 information. Levels of trust were also higher for individuals’ most trusted media source than for President Trump regardless of religious beliefs and affiliations. We also show how both religious affiliations and trust mattered, as we present evidence highlighting how Catholics, people trusting Donald Trump, and other groups were more likely to underestimate the severity of the pandemic.
Religious Americans and the News Media
Though scholars continually debate media effects on political phenomena, there is evidence that when it comes to policy knowledge, the coverage and salience of news increases individual learning (Barabas and Jerit 2009). Regarding health-related policies, there are differences depending on where individuals get their news. For example, in a study of Zika virus coverage, Jerit et al. (2019) found that local news (Tampa Bay Times) printed audience-specific, in-depth prevention coverage, compared to the New York Times. But both papers similarly used “sensationalist language and imprecise risk information” (Jerit et al. 2019, 1).
For religious Americans, particularly conservative Christians, selecting a news source for policy information can be influenced by perceptions of media bias toward their beliefs and groups. Indeed, those high in religiosity rate broadcast, print, and internet news as more untrustworthy, as compared to their less religious peers (Golan and Day 2010). Stefanone, Vollmer, and Covert (2019) showed the higher one’s religiosity, the more one found fake news credible and the more likely one would share the information.
Moreover, both political and religious elites often signal distrust in mainstream media sources, dissuading their followers from tuning in (Gaskins 2019). High religious service attenders and self-identified evangelicals consider the media hostile to their beliefs and that the media are often factually inaccurate in their stories (see also Ladd 2012). These behaviors are associated with lower political knowledge for highly religious individuals, though it is important to note like many political knowledge scales taken from large publicly available surveys, these items focus on major political figures and institutions, factors that are often less relevant to effective political decision-making (Gaskins 2019; Lupia 2016).
The scholarship on understanding when and how conservative Christians or evangelicals seek or avoid political content has not considered these effects in light of a major health crisis, a time when trust in news sources is directly related to health outcomes (Young and Bleakley 2020). With respect to source trust itself, Krause et al. (2019) found that when asked about the kinds of people they generally trust to “tell the truth,” scientists are one of the most trusted groups, even when thinking about scientific issues that can be polarizing. Yet Cacciatore et al. (2018) found that evangelical Christians use media coverage about science differently than non-evangelicals and are less likely to trust scientific sources of information.
The distrust of the news media sown by many religious elites and partisan political leaders is connected to various religious groups’ media attitudes. Mormons feel the mainstream media are hostile to their beliefs and politically biased and therefore consider these sources less trustworthy and credible (Golan and Baker 2012). The relationship with religiosity and media trust is more mixed (Golan and Day 2010). Evangelical Christians are more trusting of religious authority figures to tell the truth about scientific information, though they are only slightly less likely than non-evangelicals to trust scientific authorities for the same information (Cacciatore et al. 2018). Evangelicals also tend to use mediated information differently than non-evangelical Christians.
Certain religious beliefs may structure media preferences as well. For example, Christians who hold literal views of the Bible also may be attracted to information sources that display authoritarian themes (for a discussion, see Djupe and Burge 2021). Those with these literal and dogmatic orientations may be more likely to look to the Trump administration or Fox News for their pandemic guidance rather than a scientific, deliberative body like the Centers for Disease Control.
Modality (Talk Radio, TV, Online, Newspaper)
Given the drastic growth of media choices and the rise of niche markets in the contemporary information environment (Stroud 2011), it is imperative to update our understanding of how individuals’ media diet can be shaped by their religious beliefs. Research shows that today’s audiences structure their media repertoires based on both ideology and modality (talk radio, TV, online, print) (Edgerly 2015). While social media and online news websites increasingly outpace print newspapers in attracting audiences, radio and TV—despite being “older” media than online platforms—still take the lion’s share of Americans’ media diets (Shearer 2018). In 2018, more than 80 percent of Americans aged twelve or older listened to AM/FM radio in a given week (Pew Research Center 2019). In particular, news/talk formats—where such conservative commentators as Rush Limbaugh and Sean Hannity are prominently featured—are highly influential in the public political discourse (Sobieraj and Berry 2011). Conservative talk radio and Christian radio constitute a substantial proportion of a conservative media diet together with Fox News and conservative blogs (Edgerly 2015).
For our purposes, much of the work on religiosity and news modality preferences has not been updated (see Golan and Day 2010 for a review). For example, there were positive relationships between high religious service attenders and newspaper readership (Finnegan and Viswanath 1988; Sobal and Jackson-Beeck 1981), and religious people eschew the internet in favor of news from print and television/radio (Armfield and Holbert 2003). In 2008 and 2012 national surveys, highly religious Americans were more likely than the less religious to listen to radio for news but no more likely to tune into Fox News once partisanship was taken into account (Gaskins 2019). Part of this chapter’s contribution is to update the connections between religious belief and media consumption.
Religion and Trust
Though news and social media are important distributors of policy-related information and knowledge, government agencies also have their own messaging strategies and channels. This is particularly the case with a public health crisis like the COVID-19 pandemic. Individuals are bombarded with messages from the Centers for Disease Control, their state and local public health departments, university medical centers, clinics and hospitals, and local, state, and federal public officials. But reception of those messages hinges on trust—perhaps a generalized trust of others or a specific trust in government entities.
The literature on religious individuals and political and social trust is mixed (see Hsiung and Djupe 2019 for a discussion). While some scholars argue that political trust should be separately understood from social trust (Newton 1999), it is not clear in the case of information about a pandemic whether this trust should be classified as political or not. In fact, the politicization of COVID-19 has been the topic of much discussion—a phenomenon that should be about health and prevention has become polarized by belief and suspicion.
Hsiung and Djupe (2019) argue that variation in social and political trust cannot be neatly organized into the believing, behaving, and belonging categories often describing religion and politics in the United States (Layman 2001). They point to evidence from Traunmüller (2011), which suggests trust increases when religious people feel they are in the majority. In the case of evangelical Christians and the Trump administration, we can expect they will put their trust in Republican-affiliated government officials and only media where they feel part of this majority (e.g., Fox News). Trust in government closely follows partisan control, but also white evangelical Christians were overall less trusting of government (Hsiung and Djupe 2019). More recent evidence from the Pew Research Center (2020) shows that white evangelical Christians hold more positive views of President Trump and are more agreeable with President Trump on important issues, compared to other religious groups and those who are unaffiliated.
With respect to COVID-19, DeFranza et al. (2020) find that religiosity is associated with a decreased likelihood in following shelter-in-place orders as the orders are interpreted as a violation of personal and religious freedom, topics regularly raised in conservative media ecologies (Jamieson and Cappella 2010). Further, trust in these sources tends to be asymmetric, with conservatives trusting talk radio and Fox News at significantly higher levels than moderates and liberals (Pew Research Center 2020d).
Expectations and Open Questions
Given the strong relationship between political and religious identity (Margolis 2018) and the increasing likelihood that religious traditionalists identify as Republicans, we expect evangelical Christians to be more likely to trust the Republican president, Donald Trump, as a source of information about COVID-19. It is less clear what to expect regarding non-evangelical Christians, given the diversity of their partisanship and media use. We also expect that those with more literal views of the Bible will express more trust in President Trump to tell them the truth about COVID-19. Because evangelicals tend to interpret scientific information differently than others (Cacciatore et al. 2018), exhibiting less trust in scientific authority, we expect evangelicals to be less likely than others to trust the CDC to tell them the truth about COVID-19. We do not expect there to be differences in trust of one’s most trusted news source based on religious identity or affiliation, although it will be important to compare differences in which sources are trusted the most by religious groups.
Sample and Methods
In April 2020, we conducted a survey of U.S. adult respondents recruited by LHK Partners, with one-third of participants from counties with zero local newspapers, one-third from counties with one local newspaper, and one-third from counties with more than one local newspaper (n = 2,063). These three groups of counties were prematched with a genetic matching algorithm to ensure that they share similar demographic and geographic characteristics despite differing in the number of local newspapers they have. The sampling procedure was developed for another purpose, but the size and makeup of the sample can provide useful insights about how various religious groups approach trusting sources to tell them the truth about important public health issues.
Each respondent was asked to rate how often they access twenty-five different news sources—from talk radio to blogs to television shows—and how much they trust each source. We also asked participants how much they trust the following to “tell you the truth about COVID-19”: President Trump, the U.S. Centers for Disease Control and Prevention, and the news source the respondents had indicated that they trusted the most (a source they indicated to us earlier in the survey from a list of twenty-five sources we provided).
Measures
Religious groups. We categorized religious groups using two survey questions. The first question asked respondents whether they would describe themselves as a born-again or evangelical Christian or not, and the second question asked respondents to choose their religious preference from Protestant, Roman Catholic, Jewish, Muslim, Mormon, other Christian religion, other non-Christian religion, no religion, and don’t know. Those who answered yes to the first question and chose Protestant, other Christian religion, or don’t know for the second question were coded as “evangelical Christians” (26.27 percent). Those who answered no or don’t know to the first question and chose Protestant or other Christian religion for the second question were coded as “non-evangelical Protestants” (22.69 percent). Those who chose Roman Catholic were coded as “Catholics” (19.78 percent). Due to small sample sizes, we collapsed Jewish, Muslim, Mormon, and other non-Christian religion to create the “other religion” category (8.48 percent) and collapsed no religion and don’t know (who also did not identify as evangelical) to create the “nones” category (22.78 percent).
Biblical literalism. Respondents answered the question “Which of these statements comes closest to describing your feelings about the Bible?” by choosing from “the Bible is the actual word of God and is to be taken literally, word for word” (24.48 percent, coded as 3), “the Bible is the inspired word of God but not everything in it should be taken literally, word for word” (50.41 percent, coded as 2), and “the Bible is an ancient book of fables, legends, history, and moral precepts recorded by men” (25.11 percent, coded as 1).
Most trusted news sources. Respondents answered the question “Of all of the news sources listed, which one do you trust the most?” by selecting one answer from the list of twenty-five items (see Figure 10.1).
Trust in institutions or person to tell the truth about COVID-19. On a 5-point scale, respondents answered the question “How much do you trust the following people and institutions to tell you the truth about COVID-19?” for President Trump, the CDC, and their answer to the most trusted news source question. Their answers were recoded so that 1 equals completely distrust and 5 equals trust completely (MTrump = 2.71; SDTrump = 1.49; MCDC = 3.77; SDCDC = 0.99; MNews = 3.72; SDNews = 0.92).
Estimates of confirmed COVID-19 cases in the United States. Based on the COVID-19 Data Repository by the Center for Systems Science and Engineering at Johns Hopkins University, we calculated the average number of confirmed COVID-19 cases in the United States during the period of our survey. We then computed an index for each respondent by dividing their estimate of cases by the documented number of cases. Using one-third quantiles of the distribution of this index, we created three categories to describe the accuracy of a respondent’s estimate. If a respondent’s index fell under the one-third quantile, they were categorized as giving an “underestimate.” If a respondent’s index fell between the one-third and two-thirds quantile, they were categorized as giving an “about right estimate.” If a respondent’s index fell above the two-thirds quantile, they were categorized as giving an “overestimate.”
Control variables. We asked respondents to report their partisan identification and recoded the items to a 7-point scale (1 = strong Democrat; 7 = strong Republican; M = 4.23; SD = 2.11). Respondents also reported their political interest (1 = Follow politics hardly at all; 4 = Follow politics most of the time; M = 3.29; SD = 0.90), age (M = 52.02; SD = 16.94), gender (55.12 percent female), education (5-point scale; M = 3.10; SD = 1.43; 80.85 percent with at least some college), race (17.30 percent nonwhite), and household income (7-point scale; M = 3.26; SD = 1.66; 57 percent below $75,000).
Results
We asked our participants to identify, of the twenty-five information sources we provided, the one that they trusted the most. For evangelical Christians, television is king. Figure 10.1 shows that Fox News was the most trusted option by far, followed by local television news and national network television news. CNN, national newspapers, local newspapers, and conservative talk radio were also chosen in moderately high numbers. News content on social media, entertainment and late night programming, daytime talk, and various political blogs were chosen least often as the most trusted news source for evangelicals.
Turning our attention to non-evangelical Protestants, the diversity of media choice is notable, though once again television sources are the most likely to be named as one’s most trusted source. Figure 10.1 also shows that mainstream network television news, the third most popular choice for evangelical Christians, came out on top for non-evangelical Protestants. While Fox News was the most trusted choice for evangelical Christians and the third most trusted among non-evangelical Protestants, non-evangelicals also exhibited notable trust in MSNBC, highlighting the ideological diversity of news content preferred by non-evangelical Protestants. Local television news was the second most popular source. CNN, National Public Radio (NPR), and local and national newspapers were also commonly selected as trusted sources of news.
Figure 10.1 The Most Trusted News Sources for Evangelical Christians, Non-evangelical Protestants, Catholics, and Nones (Source: April 2020 survey.)
Similar to non-evangelical Protestants, Catholics also demonstrate a relatively diverse pattern of news trust as compared to evangelicals. Local television news was the most trusted among Catholics, followed by mainstream network television news as a close second. Interestingly, FOX and CNN came in as the third and fourth most popular sources, reflecting the ideological diversity and the strong dominance of television sources in news trust among Catholics. The nones’ most trusted news sources reflect even more diversity of information outlets. National mainstream network news was the most trusted source, but national newspapers were close behind, as were local television news, CNN, Fox News, and NPR. Nones also selected international news sites, MSNBC, and internet news aggregators more often than evangelicals, non-evangelical Protestants, and Catholics.
Given the considerable demographic differences regarding typical evangelical Christians as compared, for example, to the average none, we want to more systematically understand whether religious orientations or demographics are more central explanations of correlates of trust in various sources of information about COVID-19. Focusing on whether one identifies as an evangelical Christian or not, Table 10.1 provides estimates for correlates of trust in Donald Trump, the CDC, and one’s most trusted news source to tell individuals the truth about the COVID-19 pandemic. The first column estimates factors associated with trust in President Trump to tell the truth about COVID-19. As we expected, evangelical Christians were significantly more likely to trust Trump than others, and, of course, Republicans were more likely than Democrats to trust the Republican president. Those with more education were less likely to trust the president to tell the truth about the pandemic. No other variables were significantly different from zero in the model.
Moving to the middle column and factors associated with trust in the CDC, being an evangelical Christian was negatively signed, as expected, but was not statistically significant. Republicans and respondents who were not white were less likely to trust the CDC. Those with higher incomes and more education were more likely to trust the CDC’s COVID-19 statements. In the right-hand column, our analysis reveals that older individuals, men, Democrats, and those interested in politics were more likely to trust their most trusted news source to tell them the truth about COVID-19.
In addition to comparing evangelical Christians to others, we also break down our results by more specific religious traditions. Figure 10.2 suggests a notable difference between the level of trust evangelicals place in Donald Trump as compared to non-evangelical Protestants, who are less trusting in the former president. Evangelical Christians trust Trump to tell the truth about COVID-19 more than any other Christian group. The only group that reported higher levels of trust were Muslim respondents, though this may be an artifact (note the wider error bars) of the small number of such respondents in our data.
When we examine average level of trust members of different religious groups placed in their most trusted news source, there is not meaningful variation across religious traditions. Notably, evangelical Christians were more likely to trust Donald Trump with pandemic-related information, as compared to everyone else in our sample, and all of the religious groups (including evangelicals) we examined placed more trust in both the CDC and their most trusted news source than Donald Trump.
Because religious beliefs are correlated with authority given to sources of information (Djupe and Burge 2021), we wanted to compare how different beliefs about the Bible were related to trust in what President Trump, the CDC, and what one’s most trusted media source had to say about the virus. In Table 10.2, the left-hand column examines correlates of trust in Donald Trump to tell the truth about COVID-19. Holding a more literal view of the Bible is positively associated with trusting the then president to tell the truth about COVID-19. When controlling for biblical literalism, only Catholics were significantly more likely to trust Trump as compared to nones, who were the baseline comparison group. The coefficients for evangelical Christians and non-evangelical Protestants were positive but not significant.
Figure 10.2 Trust in Trump to Tell the Truth About COVID-19 by Religious Groups (Source: April 2020 survey.)
The middle column shows that non-evangelical Protestants were more likely to trust the CDC to tell the truth about COVID-19 while evangelicals, Catholics, and those of other religions were not different from zero. Holding other factors constant, biblical literalism has no significant effect on trust in the CDC. Democrats were more likely to trust the CDC while those who are not white were less likely to trust the government organization. Those with more education and higher incomes were also more likely to trust the CDC.
The final column reveals that biblical literalism is positively associated with trust in one’s most trusted news source about COVID-19. Similar to findings on trust in Trump, the effect of religious traditions is muted when taking account of biblical literalism. Holding other factors constant, evangelicals, non-evangelicals, and Catholics were no more likely than nones to trust their most trusted news source to tell the truth about COVID-19. Democrats were more likely to trust in their most trusted news source as were those who were older, had higher levels of interest in politics, and identified as men.
We also examine predictors of the consequences of trusting different sources for information about COVID-19 (tables and figures included in the online appendix). We used the number of COVID-19 cases in the United States at the time of our interview with them as the baseline of comparison. We placed responses into one of three categories: “overestimating” the number of COVID-19 cases (respondent’s answer/real number falls above the two-thirds quantile), “underestimating” the number of COVID-19 cases (respondent’s answer/real number falls under one-third quantile), and “about right” answers (respondent’s answer/real number falls between the one-third and two-thirds quantiles).
More interesting are the factors associated with underestimating the number of COVID-19 cases in the United States. Trust in Donald Trump as a source about the pandemic was positively related to underestimating COVID-19 cases. However, having a high level of trust in one’s most trusted news source was correlated with being less likely to underestimate the number of COVID-19 cases; that is, news trust makes people more accurate in their COVID-19 estimates. Catholics were likely to give wrong answers, either overestimating or underestimating the number of COVID-19 cases, rather than giving an about right answer. Finally, evangelical Christians and those identifying with other religions apart from Christianity were more likely to underestimate COVID-19 cases.
Conclusion
Trust in scientific information—on topics ranging from climate change to genetically modified foods to best practices to combat a global pandemic—can serve as a heuristic for people when they consider information they encounter about science (Brossard and Nisbet 2007; Hmielowski et al. 2014). In general, and consistent with past research, the CDC was the most trusted entity when it came to people’s evaluations of who would tell the truth about COVID-19 (Krause et al. 2019). Notably, and in conflict with past research, evangelical Christians were not significantly less likely to trust scientific sources like the CDC as compared to other individuals.
People’s most trusted news sources were as likely, and in some cases were more likely, to be trusted than the CDC to share accurate pandemic information. If people’s most trusted news sources were sharing the verifiable truth with their audiences, this would be little cause for concern. However, evangelical Christians’ most trusted news source was Fox News. Increased Fox viewership was associated with a decreased likelihood of staying home in the early days of the pandemic (Simonov et al. 2020). Fox News viewers were also less likely to prepare for the pandemic by purchasing materials that promoted good hygiene and engaging in behaviors that would decrease transmission of the virus (Ash et al. 2020). Thinking about these results in concert with our own—including our demonstration that literal believers in the Bible were more likely to trust their more favored news source on COVID-19—it is possible that evangelical Christians who trusted Fox News and other sources sharing information that played down the seriousness of COVID-19 may have been more likely to avoid pro-social health behaviors or even more likely to become infected with the virus.
Though their faith in their most trusted news source was higher, evangelical Christians, non-evangelical Protestants, and Catholics all were more likely than nones and people of other religious backgrounds to believe that Donald Trump would tell them the truth about COVID-19. The same was true of both Christians who said the Bible should be taken literally and Christians who expressed belief that the Bible was the inspired word of God. This is consistent with the demonstration that religious traditionalists across denominational perspectives are more likely to support Republicans (Layman 2001).
Still, there is considerable heterogeneity in how people of different faith perspectives and varying levels of fervency in their religious beliefs apply their religious lives to their political views (Friesen and Wagner 2012). For those whose religious identities match their political identities, we might expect even higher levels of trust in co-partisan sources of scientific information than those whose religious identity does not perfectly overlap their partisan one. It is important for future research to understand religion’s role in how political leaders’ rhetoric can cement polarized partisan identities for many while excluding others from a political system that does not represent them (Wagner and Friesen 2021). That is, as scholars continue to demonstrate the political consequences of those who have overlapping identities as compared to those who do not (see Davis and Mason 2016), our evidence suggests that taking account of people’s religious belonging and believing are important for creating a more comprehensive accounting of public opinion and political behavior in the United States.
It is also important to note that joining a new faith community or leaving religion behind is often driven by local factors. This is especially true for evangelicals who place a premium on the social and cultural capital they build with fellow members of their congregations (Djupe et al. 2018). Just as polarized attitudes are driven, in part, by local context (Suk et al. 2020), knowing how congregants came to join their current flock (whether they were denomination shopping, invited by a member of their social network, etc.) is critical to placing their religious identity in context. In other words, just as it is the case for trusted news sources, religion can act as an information broker.
To help unpack observational equivalence between the level of trust most of our respondents expressed in the CDC and in their most trusted news source, future research designs might seek to force a choice between different sources of information rather than discretely evaluating trust in each. It would also be helpful to explore a richer set of questions related to religious belief, networks, and context to understand how individuals’ various identities—religious, political, and the like—interact with their spiritual, political, and social beliefs. Finally, larger samples of groups like Muslim Americans and Jewish Americans, for instance, are needed to understand how religious faith more fully is associated with believing varying sources of scientific and political information.
____________
Material referencing an appendix in this chapter can be found online available here: https://dataverse.harvard.edu/dataverse/epidemic_among_my_people.