______
Is the Effect of Religion “Raced” on Pandemic Attitudes and Behaviors?
ANGEL SAAVEDRA CISNEROS, NATASHA ALTEMA MCNEELY, AND PAUL A. DJUPE
Though the novel coronavirus pandemic has spared no group, racial and ethnic minority communities have been disproportionately affected due to existing systemic inequalities like unequal access to health care and personal health status inequalities (Alimi et al. 2020; Substance Abuse and Mental Health Services Administration 2020; see also Chapter 5). Existing work has documented that Hispanics have been disproportionately affected by COVID-19 throughout the United States and especially in Texas (Andres-Henao and Crary 2020; Villarreal 2020). Black communities, especially, have been disproportionately affected by COVID-19 (Boorstein 2020; Moore et al. 2020; Wright and Merritt 2020), bearing a high rate of infection and fatalities. And Asian communities in the United States have faced tremendous prejudice in reaction to former president Trump’s labeling of the coronavirus the “Asian flu” and a Chinese conspiracy against him and the world.
In the face of such inequities, community leaders often play an important role exposing the nature of problems, calling for their redress, and guiding their communities. In the Black community, congregations and religious leaders have performed this service, in part due to the oppression that limited the development of Black leadership in other institutions (e.g., Lincoln and Mamiya 1989). These dynamics may not be limited to the Black church, of course, since all congregations are formed within communities with particular sets of concerns. At the same time, congregations are not just neutral in their outreach and seek to impart worldviews that guide thought and action (Leege and Kellstedt 1993), some of which may be opposed to engaging with the world. That is, we cannot assume that all congregations promote robust individual citizenship, trust in governmental institutions, and social action (e.g., McRoberts 2003). Some develop a reliance on the church and faith as sources of healing and protection (see Chapter 1), some encourage social action, and some pursue other ends. This leads to a puzzle regarding the coronavirus pandemic. If religion is particularly important in minority communities, did social distancing and other public health measures undermine a key source of community influence? Was religion able to rise to the occasion to champion public health measures equally across racial groups, or did it work against public health measures as competitors with religious worldviews?
Religion Is Raced?
As a result of the racial animus that has been at the heart of American politics since the beginning, religious institutions have often developed separately across communities formed largely along racial and ethnic lines. Christianity was imposed by white slaveholders, but congregations became incredibly important resources for the Black community under slavery and after it (e.g., Lincoln and Mamiya 1989). White oppression imposed a political leadership and civil society vacuum that the Black church swelled to fill. Therefore, it is no surprise to find much political leadership in the Black community with various ties to the church. Given the heavy concentration of Latinos in the Catholic Church, we do not find the same tight connection of church and political leadership, though the Catholic Church has been helpful in knitting the community together (Djupe and Neiheisel 2012; Jones-Correa and Leal 2001). Nevertheless, urban religious organizations play an important social and political role for Latinos. It is notable that involvement in Protestant congregations does not have the same effect on Latinos as in Catholic parishes (DeSipio 2007), which may point to the power of a link to the community.
Perhaps because religious institutions have been at the core of social support, it has been widely documented that racial minorities are more religious than white Americans on most measures. The Pew Religious Landscape Study finds that while 49 percent of white Americans report that religion is very important in their lives, 59 percent of Latinos and 75 percent of African Americans report this to be true, though only 36 percent of Asians do (Pew Research Center 2014). Across a variety of measures, Black Americans report the highest levels of religious behaviors, and yet their political attitudes and behaviors differ in important ways from those of other religious Americans. Even so, we cannot assume anymore that racial/ethnic minorities are attending racially homogeneous congregations. As Pew Research Center’s (2021) Faith among Black Americans report makes abundantly clear, historically Black churches command a minority of Black membership and may not even be the plurality among younger African Americans. Even greater numbers of Latinos appear to be worshipping in multiracial congregations (Wong 2018, 20).
Since racial minorities have disparate life circumstances from whites in the United States, on balance, their religious communities are likely to take on a different set of priorities that are likely to collect under a social justice umbrella. That is, we would expect congregations with a greater number of minorities to advocate for civil rights, equality, and a more robust social welfare system. That does not mean that they will be more liberal on all issues, and many racial minorities, especially evangelical identifiers, take conservative stances on social issues like abortion and same-sex marriage (see Wong 2018 for a comprehensive look).
Some of these patterns have been affirmed by congregational data. Moreover, not all congregations become social service providers. For example, many of the connections Black congregations make with community organizations involve civil rights organizations and a sizable number do not partner with an organization (Lincoln and Mamiya 1990, 151). In turn, such partnerships are important linkages, with the congregation offering social services as well as willingness to seek government funding for those initiatives (Owens 2007). Rates of community engagement may be higher in the Black community, but congregations are not invariant.
It is no surprise that congregations take the needs of their congregations seriously, and their concern can take many forms, from the material to the symbolic. Religious organizations have historically created successful partnerships with health organizations (Solari-Twadell, Djupe, and McDermott 1990) and public schools (Galiatsatos et al. 2020). For our purposes, researchers have noted the importance of partnerships between public health and government officials with cultural as well as faith-based groups to increase compliance with COVID-19 guidelines (Alimi et al. 2020). One benefit is the mitigation of the spread of misinformation among vulnerable groups, including racial and ethnic minorities (Clark-Ginsberg and Petrun Sayers 2020). Preliminary evidence demonstrates the effectiveness of successful partnerships between public health organizations and faith-based organizations to increase compliance with COVID-19 guidelines within the African American community, while also providing needed resources to its members (Akintobi et al. 2020; Thompkins et al. 2020).
In politics, clergy tend to take greater representational roles for their congregations when they lack it locally. That is, clergy are more likely to take community leadership roles in communities with fewer active leaders (Olson 2000), and clergy step up their representation roles when their congregations are in the minority locally (Djupe and Gilbert 2003). If these dynamics exist across the religious spectrum for white clergy (Djupe, Burge, and Calfano 2016), it is highly likely that we would see it replicated in minority communities as representational roles have been documented historically (e.g., Morris 1984).
This is another way to say that there is real religious variation within minority communities. Partly due to small sample sizes of racial minorities in samples, but also because “differences across Protestant affiliations pale in comparison to structural and cultural similarities resulting from the legacy of racial discrimination and inequality” (Shelton and Cobb 2018, 737), it was common practice to identify a singular category of Black Protestants. From a religious perspective, that assumption is untenable, and there is a wealth of religious variability within the Black community. For instance, Black Methodists (e.g., the African Methodist Episcopal Church [AME]) and Black Baptists developed apart and differed in the degree of denominational organization as well as on whether clergy should be specially educated (Shelton and Cobb 2018). Moreover, Black denominations tend to differ in their public theology, with some affiliates, like Church of God in Christ (COGIC), less likely to support political activity than AME affiliates, thought to be linked to the “other worldliness” of COGIC (McDaniel 2003).
One of the key remaining questions pushes us to consider the endogeneity claim—that the politics of minority religious groups is simply a function of who sits in the pews. We wonder whether minority religion is, in itself, important to consider. Does religion work differently in minority communities? Existing research is suggestive on this point but unable to answer this question definitively because political stances are quite often overdetermined—there are many forces acting on the politics of racial minorities. That is, it is hard to set aside the American political context that is so thoroughly shaped by racial politics. Put another way, Black presidential vote choice has been almost unanimous for Democratic candidates for decades and pays no regard to variation in religious identity, belief, or behavior (McDaniel and Ellison 2008).
Studying religious influence on political attitudes often runs into this problem—many of the issues have been around for decades. So, associations between religious identities and attitudes may be the result of long-running socialization, critical events that have long passed, or other undocumented forces that happen to correlate with religion. These problems, therefore, help provide guidelines to studying religious influence: (1) we can gain greater confidence when issues are novel; (2) we need to be able to document religious input into the system (“exposure”); and (3) we need to assess whether that input was adopted to shape attitudes or behavior.
In this regard, the coronavirus pandemic may be particularly useful, especially at the beginning of its spread. As we will show, the racial disparities exposed as the pandemic progressed had not revealed themselves in late March, which was, depending on the state, about two weeks into lockdowns. The threat perceived from the pandemic was near universal, and the response from congregations was manifestly similar, if certainly not unanimous. Therefore, the early stages of the pandemic provide us with an essential baseline to assess whether religious responses varied upon widespread perceptions of threat.
The Argument
We examine how religious experiences impact COVID-19 attitudes and behaviors among non-Hispanic white, Latino, Black, and Asian Christian believers. Focusing on religious and political predictors without taking into consideration how race and ethnicity shape the American religious experience can obfuscate dynamics that are important to understanding American reactions to COVID-19. We build from the recognition that places of worship can help bridge information gaps for disadvantaged communities. That is, since minority communities have been hit hard by the pandemic, we expect that their religious organizations will pay special attention to the pandemic and provide religious justifications for taking public health seriously. We expect that those who hear their clergy engage the pandemic will assess the virus as a more serious threat and take personal and collective health seriously. This response should be particularly strong within racial minority communities, as clergy have historically been community leaders, and their messages will resonate with the problems people are facing.
Yet religion is not infinitely flexible and may not respond in the same way to community needs. That is, some religious communities are constrained by their beliefs. For instance, a growing number believe that religious belief is sufficient to ensure health, that sickness is a sign of sin, and health is a sign of godliness. This belief, often referred to as the Prosperity Gospel, is particularly common among racial minorities and white evangelicals (see Chapter 1). We suspect that belief in the Prosperity Gospel will procure a more defiant pandemic response equally across racial groups.
Data and Measurement
More and more survey efforts are taking religion seriously, but few include questions sufficient to understand what religion institutions are doing. That is clearly a pressing matter in the pandemic and central to our chapter. Fortunately, the late March 2020 survey from Djupe, Lewis, and Burge included such questions, which enable an assessment of whether religious institutions in which racial/ethnic minorities worship are responding differently than whites’ houses of worship (whites = 2,049; Blacks = 353; Latinos = 208; and Asians = 113). Critically, the survey asked respondents whether their clergy had addressed the coronavirus at all.
It is important to acknowledge that using respondent reports is not a silver bullet for capturing exposure to communication within religious institutions—it is likely to come with measurement error. As Djupe and Gilbert (2009) point out, there is rampant misperception of what clergy and congregations are doing from the viewpoint of congregants. They sometimes pay little attention, they downplay communication they disagree with, and sometimes they aren’t in the pews when a message is delivered. Still, asking congregants is at least one way in which we can get leverage on this important question.
We consider several responses to the pandemic. The survey asked whether the respondent was being encouraged to attend services in person (just who was doing the encouraging is left unspecified). It also asked whether the respondent was social distancing, which was defined as “staying home as much as possible, avoiding social contact.” And it asked about whether their congregation should defy state orders to close, should they be made.
They were in the field again in October (n = 1,790) and repeated some, but not all, of the questions asked in March. While just before the presidential election, it was also a time when COVID-19 cases were reaching stratospheric heights, far greater cases of infection than in the spring. By October, disparities across racial/ethnic groups were well established. Where we can, we compare the March numbers with those found in October.
It is not sufficient to merely examine the difference of means between racial/ethnic groups since engagement with religion differs among these groups, and the level of exposure to institutions will affect what messages get through. Therefore, we examine whether clergy were more likely to engage given respondents’ level of attendance at worship services.
Results—Clergy Engagement
Our key hypothesis is that racial/ethnic minorities will report their houses of worship addressing the pandemic at higher rates. Figure 6.1 shows this first for each racial/ethnic group and then when incorporating worship attendance rates (in the right panel). The results confirm that whites were the least likely to report hearing their clergy address the pandemic (only 33 percent). Latinos report pandemic engagement at a higher rate than whites, but the gap is not significantly larger (because of the small sample size). But both Asians (43 percent) and African Americans (46 percent) reported distinguishably higher rates than whites did. It may be surprising that it is not higher, but, again, some of this may be due to variation in worship attendance—you cannot hear a message when you’re not in the pews.
Figure 6.1 Report Their Clergy Addressing the Pandemic by Racial Group and Worship Attendance (Source: March 2020 survey.)
The right panel of Figure 6.1 shows how these reports differ when we consider worship attendance levels. Among those who attend weekly or more often, there is no difference in reports of clergy pandemic engagement among racial/ethnic groups with nearly 60 percent reporting it. The differences between groups only grow at lower rates of attendance to the point where about 20 percent more Black and Asian Americans (compared to Latinos and whites) reported clergy pandemic engagement among those who attend only a few times a year.
This pattern is suggestive of the frequency of engagement with the pandemic. For infrequent attenders to report hearing a message at nearly the same rate as frequent attenders, the message must be on heavy rotation. Without a prominent and lasting place on the clergy’s agenda, infrequent attenders would simply miss it, as appears common among Latinos and whites. But this pattern also suggests that not all religious institutions were discussing the pandemic. If they were, then the rate of clergy pandemic engagement would be much higher among persistent attenders.
By October, the rate at which clergy addressed the pandemic “in your house of worship this year” had declined across the board. Only 30 percent of African Americans and whites, 27 percent of Asians, and 35 percent of Latinos reported their clergy addressing the coronavirus. The same link to prepandemic worship attendance remained, but a combination of service cancellations and weaker engagement with online worship surely cut into hearing from clergy.
In-Person Worship
One of the key and most controversial recommendations from public health officials was to limit social gatherings in size and frequency to “flatten the curve.” As we have read about in multiple chapters in this book, these recommendations and mandates were not taken particularly well from some sectors. Religious conservatives were arguably the most incensed. The survey asked respondents whether they were being encouraged to attend worship services in person. Again, the source of this encouragement was not specified, so we cannot pin it on clergy, but we can assess whether it was more likely to be reported by frequent attenders.
Racial minorities were more likely to report encouragement to attend in person than whites, though only significantly so for Latinos—37 percent of Latinos reported encouragement, which is about 10 percent more than Black and Asian Americans (see Figure A6.1 in the appendix). A not inconsequential 23 percent of whites reported this encouragement. It is no surprise that receiving encouragement increases by typical (prepandemic) worship service attendance, though the increase is much more muted among African Americans—a 5 percent increase versus approximately 20 percent among the other groups. At best, this is modest evidence that religious effects vary by race but generally highlights that people who are deeply engaged in religious (or other) institutions will receive a greater pull to remain involved in person.
Actual reports of in-person worship in March were higher than one might expect. Upward of 90 percent of Asians, whites, and Latinos reported that in-person worship had been canceled by the time of the survey, though only 80 percent of Black Americans so indicated. From another measure, Black Americans were attending in person in late March at higher levels—25 percent—compared to 17 percent of whites and 19 percent of Asians. By October, despite the rapidly swelling caseloads, in-person worship had drastically increased so that 48 percent of Black Americans reported attending in person, which was not much higher than the 44 percent of whites, 40 percent of Latinos, and 37 percent of Asians. We do not know the extent to which in-person worship was socially distanced and operating at full capacity, but we can tell that these patterns of worship during the pandemic reflect prepandemic worship rates.
Prosperity Gospel
One of the weaknesses of the clergy pandemic engagement item, of course, is that we do not know the content of the message. Did they sound the alarm about the pandemic and encourage compliance with public health orders? Or did they throw sand on the whole project and downplay the severity of the threat posed by the virus? One way we can approximate content is by examining whether reports vary by religious beliefs. Perhaps the most important religious belief in this pandemic is the Prosperity Gospel—the belief that health and wealth on earth are the payout for fervent belief in God (see Chapter 1 for discussion and measurement details; see Chapter 2 for further results). Given the structure of this belief system, we suspect that the links with pandemic responses will not vary by race. Prosperity gospel believers will be fervent advocates of continuing in-person worship and will receive encouragement to remain so.
That is precisely what we find from a statistical model that controls for worship attendance and demographics (see Figure A6.2 in the appendix). Though Latinos report more encouragement to attend in person across the board, the three other racial/ethnic groups show steadily increasing encouragement as their Prosperity Gospel belief grows. About 40 percent of the most committed prosperity gospelers of any race report being encouraged to attend in person. That drops to close to zero for all groups except for Latinos at the lowest level of Prosperity Gospel belief.
Social Distancing
By late March, most religious organizations were complying with public health orders (or recommendations) and closed to in-person worship. By April, estimates are that 90 percent of congregations stopped offering in-person worship (PRRI 2020), though, as we see in Chapter 16, perhaps three-fifths were able to pivot to remote-access worship in some form. That does not mean that everyone was on board with this decision, and it does not necessarily imply that members took this lesson to heart in their lives outside of the congregation.
We can ask about social distancing in a survey, but we should be aware that the responses are likely to be inflated by social desirability bias—people knew it was the right thing to say. Still, there was some variation in the responses, which we chart according to prepandemic worship attendance and whether their clergy engaged with the pandemic. Figure 6.2 highlights that clergy pandemic engagement helped sustain social distancing. With the exception of Latinos, among whom attendance and clergy engagement has no effect, those attending more often where the clergy avoided talking about the pandemic showed a decreased commitment to social distancing compared to those who attend less often. However, when their clergy engaged, then commitment to distancing remained as high as it did for infrequent attenders.
That is, church attendance could be considered a measure of sociability or of the need for social interaction. Some make this case when they link religion with pro-social behavior (e.g., Saroglou et al. 2005; Shariff and Norenzayan 2007). But in the case of the pandemic, we also have the competing interpretation that attenders are being defiant about closing congregations to maintain their treasured social gatherings (see also Chapter 1). Even so, it seems clear that clergy were trying to encourage members to follow public health orders beyond the congregation since reported social distancing behavior is higher when they report their clergy engaging the pandemic. Moreover, there’s not much evidence that this effect differed by race, at least not in the sense that it reflects the severity of the pandemic for minority racial communities. For Asians and whites, clergy engagement changed the effect of attendance. Without clergy engagement, more attendance drove down social distancing. But when clergy engaged, that decline was arrested, and, for whites, the effect of attendance became slightly positive (more commitment to distancing). For Black and Latino Americans, there was a slight bump in commitment to distancing when their clergy engaged, but it was constant across attendance levels.
Figure 6.2 Does Religious Engagement and Communication Affect Social Distancing by Race? (Source: March 2020 survey.)
Defying Public Health Orders
One of the most notable stances of religious organizations during the pandemic was defiance. Though not practiced by many, there were high-profile instances of holding services despite closure orders that resulted in at least one arrest in Louisiana (e.g., Associated Press 2020b). Even if the incidence of actual defiance of health orders was evidently low (perhaps 10 percent of congregations were open and not all states required closure), the attitude that congregations should be defiant was quite a bit higher in our March sample—22 percent agreed that “If the government tells us to stop gathering in person for worship I would want my congregation to defy the order.” That figure varied by race, with racial/ethnic minorities taking stances of greater defiance—about 30 percent of minorities took the defiant stance versus 18 percent of whites. Did clergy pandemic messaging change the link between religious engagement and defiance of health orders?
As Figure 6.3 shows, the effect of more religious observance is to drive up the prospect of defiance of government public health orders, though it is important to note that none of these groups are estimated to have a majority opposing health orders. In most cases, weekly attenders show from half to a full scale point (12–25 percent) more support for defiance than those who never attend.
The effect of clergy engagement is interesting. In contrast to select media reports of clergy openly resisting closures, when clergy are reported to have engaged the pandemic, defiance drops. In most every case in Figure 6.3, support for defiance is lower when clergy engage—in the case of Blacks and whites, it is a fairly dramatic difference (about half a scale point at its maximum). Still, in all cases except for Asians, more attendance drives up a defiant stance even when clergy engage. That either means the content of what clergy are urging is different in high-attendance churches or clergy are unable to arrest this sentiment among their most faithful members.
Figure 6.3 Does Religious Engagement and Communication Affect Defiance of Public Health Orders by Race? (Source: March 2020 survey.)
Notably, this pattern changed by October 2020 (results not shown). For Asian Americans, worship attendance is linked to a less defiant stance, as it is among Latinos whose clergy talked about the pandemic. Attendance has no effect among whites but has opposite effects among Black Americans depending on clergy engagement. African Americans whose clergy talked about the pandemic provided a more defiant stance against public health orders, while attendance is linked to a less defiant stance without clergy engagement. It seems clear that the community circumstances were not the same in October across racial groups.
Conclusion
One unequivocal lesson from the 2020 pandemic experience is that it hit minority communities harder than white ones, though the virus reached every corner of the United States and the world. As a result, it is natural to suspect that religious organizations would reflect this distribution and engage with the pandemic in different orders of magnitude. Religious organizations have been central to racial/ethnic minority civil society in the United States, demonstrated still by higher levels of religiosity. This creates an analytical problem, however. Does diversity in the pandemic response among racial/ethnic groups reflect the community, in which the religious organization happens to be apart? Or does the local religious organization have a value-added response to the pandemic?
Our tack was to first check whether congregations engaged with the pandemic, which seems to be a necessary condition for asserting their effect (see, e.g., Wald, Owen, and Hill 1988). In March 2020, only a minority indicated that their clergy spoke about the pandemic, though it was higher among Black and Asian Americans, which is what we would expect. That level, though, was purely a function of religiosity levels. Then we tested what kind of pandemic response was linked to that communication. We expected that religiosity would be related to different responses by race that reflected the seriousness of the problem in the community but found little evidence of it.
Weekly attenders of all racial/ethnic groups reported the same level of clergy engagement. There were gaps at lower levels of attendance, however. A sizable minority reported being encouraged to attend in person, which violates the conventional wisdom that social distancing would help limit the spread of the virus. Again, there were few differences among frequent attenders of any race, though Latinos reported the highest rates of encouragement to attend in person. One thing that appears to help unify responses across racial groups is adherence to the Prosperity Gospel. Higher adherence to that worldview is associated with greater encouragement to worship in person and about equally among racial/ethnic groups.
This set of results brings us in line with the findings from Khari Brown and colleagues’ investigations of immigration attitude variation by race and exposure to cues in churches (Brown 2010; Brown and Brown 2017; Brown et al. 2017). They find, in broad brush, that the efficacy of clergy communication hinges on different measures of threat, such as financial precarity. In our study, different racial groups reported almost identical levels of threat from the coronavirus, varying less than a tenth of a point from each other on a 1–5 scale. With that knowledge, it is perhaps unsurprising that at this early stage of the pandemic, religious cues had roughly the same effects across groups.
Moreover, clergy communication appears to be largely in sync across groups. We can only infer the content of that communication because it was not asked in the survey. But the patterns are relatively consistent that clergy are consistent promoters of public health messages. That does not mean that everyone was on board with closures and other public health measures, but respondents were more cooperative when clergy engaged—expressing more commitment to social distancing and less defiance to state orders to stop holding in-person worship.
This chapter continues the conversation about whether and how religion in the United States is raced (e.g., Brown et al. 2017; McKenzie and Rouse 2013). It highlights the analytical importance of having some measure of exposure—here, explicit engagement with the pandemic by clergy. Once we have that and can account for the differences in religious engagement across racial/ethnic groups, then we find little evidence that congregational responses across the United States varied by race. Of course, this does not mean that this same pattern will be found in other issue areas. But this is arguably a strong signal given the immediate and strong impact the pandemic has been having on minority communities.
____________
Material referencing an appendix in this chapter can be found online available here: https://dataverse.harvard.edu/dataverse/epidemic_among_my_people.