15

______

How the Early Stages of the COVID-19 Pandemic Affected Religious Practices in the United States

KRAIG BEYERLEIN¹ AND JASON KLOCEK

On the morning of Sunday, September 16, 2001, Pastor Tim Keller arrived to a line of congregants out the door of Redeemer Presbyterian Church in Manhattan. The typical crowd of about 2,800 people nearly doubled the first weekend after the September 11 attacks. In fact, so many people showed up that Pastor Keller had to tell those who could not fit inside the church to come back in two hours for another impromptu service. Some 800 people returned (Zylstra 2017).

The spike in turnout at Redeemer Presbyterian Church was neither unique to that congregation nor to New York City. According to a 2001 Gallup poll, religious service attendance rose 6 percent across the United States the Sunday following 9/11 (Walsh 2001). The Wall Street Journal reported on a study that found a 25 percent increase across religious congregations and spiritual centers immediately after the attacks (McLaughlin 2001). Regardless of how one measured the initial surge in religious activity, it also proved to be remarkably short-lived. In just two weeks, most U.S. congregations reported that attendance had returned to pre-9/11 levels (Iannaccone and Everton 2004).

The abrupt and fleeting rush to congregations following the September 11 attacks raises important questions about the impact of large-scale catastrophes—such as famines, wars, earthquakes, and pandemics—on religiosity. Yet our ability to test this relationship and draw meaningful conclusions remains limited because we typically lack comparative data from before and after disasters. As such, we still know relatively little about the nature, size, and duration of large-scale tragedies’ effects on religious and spiritual practices.

Just as importantly, the empirical study of religious behaviors after wide-reaching crises relies on a relatively limited set of cases. Most existing research focuses on natural disasters that affect a narrow geographic region. The nature of the COVID-19 pandemic, however, introduces unprecedented dynamics. What happens when the majority of congregations in a nation are all suddenly closed or inaccessible? How do people respond when important in-person religious practices they would normally turn to for solace during troubled times pose a significant health risk to themselves and others? Do they replace these practices with safer forms or continue to gather in person and hope they will not be infected?

This chapter draws on representative survey data from U.S. adults collected six weeks into the COVID-19 pandemic to answer these and related questions. We leverage measures asking about religious behaviors in 2019 and six weeks into the coronavirus outbreak among the same respondents to examine whether they increased, decreased, or remained the same. Our over-time analyses focus on four specific religious practices: attending in-person congregational services, viewing virtual religious services, gathering together in person for informal religious activity, and praying privately. Moreover, these analyses are conducted at two distinct levels: (1) wave to examine overall change in these practices and (2) respondent to identify stability or movement in them among individuals.

The remainder of the chapter proceeds in four parts. In the next section, we summarize extant scholarship on religious responses to crises. Much of that research focuses on how personal trauma, rather than large-scale catastrophes, affects religiosity. The subsequent two sections describe our survey data/methods and findings, respectively. Several wave-level changes in religious practices, including a sharp decline in attending religious services in person and an uptick in viewing virtual religious services during the initial coronavirus outbreak, are in line with theoretical expectations. Others, such as the general stability in private prayer frequency, are more surprising. The final section discusses the implications of our results and identifies areas for future research.

Existing Research on Religion and Crises

Social scientific research on the role of religion in how people experience and respond to crises has developed considerably over the last several decades. Psychologists have carried out the bulk of that work, concentrating on how religious beliefs and practices facilitate coping with illness, the death of a loved one, sexual and physical abuse, or other personal traumas (e.g., see Pargament 1997; Pargament and Raiya 2007). This research has uncovered a range of positive and negative religious methods that people employ to understand and manage unexpected and distressing life events (e.g., see Luhrmann 2013; Pargament 1997; Pargament and Raiya 2007). A considerable amount of attention has also been given to testing empirically the impact of these methods on physical and mental health outcomes (e.g., see Abu-Raiya et al. 2020; Pirutinsky, Cherniak, and Rosmarin 2020). The emerging consensus from this scholarship is that reliance on positive religious coping methods generally has salutary effects in the face of individual trauma.

In the last two decades, natural disasters that strike specific communities have become another area of interest for studying the impact of crises on religious attitudes and behaviors (Belloc, Drago, and Galbiati 2016; Sibley and Bulbulia 2012). Much of the attention has focused on meteorological events, such as floods, droughts, hurricanes, and wildfires. Several studies, for instance, have observed an increase in religious practices (e.g., praying, Bible reading, attending services in congregations) among African Americans living in communities that Hurricane Katrina hit (Alawiyah et al. 2011; Chan and Rhodes 2013). Others have noted a general increase in survivors’ self-reported levels of religiosity following the devastation in the U.S. Gulf Coast region in 2005 as well as after the 1993 Mississippi River floods across the Midwest (Kessler et al. 2006; Smith et al. 2000). And one study of Canadians who endured weather-related disasters with significant fatalities found an increase in both religious belief and behavior among respondents (Zapata 2018).

Geological disasters, especially earthquakes, are another focus in the literature regarding how tragedy influences religiosity (e.g., Jang, Ko, and Kim 2018; Stratta et al. 2012). One recent cross-national analysis, for instance, found that individuals in regions with higher earthquake risk tend to be more religious (e.g., believe that God is very important in their lives or attend religious services at higher levels) than those living in lower-risk areas (Bentzen 2019). Another study of the 2011 earthquake in Christchurch, New Zealand, observed that religious affiliation increased in the devastated region, while the rest of the country experienced a net decrease (Sibley and Bulbulia 2012).

Social scientists have paid less attention to religious responses to present-day catastrophes that affect a large segment of the population, a whole nation, or the entire world. This is partly because events of this magnitude remain relatively rare. These events are no less important to understand, and possibly more so, given they have far-reaching and lasting repercussions. As discussed at the beginning of this chapter, 9/11 has been the primary case investigated to understand how contemporary large-scale tragedies influence U.S. religious attitudes and behaviors. This includes a handful of analysts who have documented a brief uptick in religious service attendance (Iannaccone and Everton 2004; Uecker 2008) and others who have shown an increase in self-reported measures of religiosity and spirituality (Bonanno and Jost 2006; Seirmarco et al. 2012; Uecker 2008) after the September 11 attacks.

The COVID-19 pandemic constitutes a new opportunity for scholars to explore how large-scale crises influence religious beliefs and behaviors given its unprecedented dynamics, including the extended period over which the tragedy has unfolded. Additionally, the coronavirus outbreak has involved unique restrictions on in-person access to religious spaces. During the initial weeks of the pandemic in the United States, large numbers of congregations voluntarily suspended services out of an abundance of caution (Mervosh and Dias 2020). The vast majority of them that did not close initially eventually did so in response to states’ stay-at-home orders. And even when those constraints were loosened or lifted, many faith communities continued to hold in-person services or placed limits on the number of people who could gather together (Dias 2020). In response, faith leaders significantly expanded online religious services (Estrin 2020). These circumstances raise additional questions. For instance, did those who previously attended in-person religious services seamlessly transition to a virtual platform? Or did they find online services too impersonal? What about the possibility of attracting new people? Will the greater accessibility of streaming services lead those who previously did not attend in-person services to join online more regularly?

Developing research efforts attempt to document and explain changes in religious behaviors in response to the unusual circumstances of the COVID-19 pandemic. Outside of the United States, one study, for instance, found an increase in online searches about livestreamed Catholic masses in Italy during the initial lockdown there (Alfano, Ercolano, and Vecchione 2020). Another observed that Italians who reported a case of coronavirus in their household also reported higher levels of attendance at religious services and frequency of prayer compared to those who reported no household infection (Molteni et al. 2020). Still another recent analysis identified a global rise in Google searches about prayer since March 2020 (Bentzen 2020). Additionally, two recent Pew Research Center studies provide some insight into how the coronavirus outbreak has affected religiosity in the United States. One conducted in March 2020 documents a shift from in-person to online religious services at the very beginning of the outbreak (Pew Research Center 2020b). Another in July 2020 reports that more than half of U.S. adults said they plan to resume going to religious services as often or more often as they did before the pandemic once it is safe to do so (Pew Research Center 2020a).

This nascent research provides a useful starting point for thinking about the impact of the COVID-19 pandemic on religious practices. Most notably, it points to a general expectation of their increase in response to the current crisis. But this research is not without data challenges. One is that collected data are generally for online searches about various religious practices (i.e., prayer and virtual services) rather than people engaging in them. We should not necessarily assume that these are proxies for religious behaviors in real life. In fact, our findings cast doubt that they are. Another limitation is that existing studies draw overwhelmingly on aggregate measures of religious behaviors (at country or regional levels). This limits our ability to infer whether the observed relationships hold at the individual level as well as to determine whether any observed difference is primarily the result of people changing the ways in which they practice their faith or new people engaging in religious practices for the first time. Furthermore, much of the data analyzed thus far come from outside the U.S. context and do not directly compare religious behaviors before and during the COVID-19 pandemic.

Our chapter addresses a number of these concerns. Most importantly, we draw on survey data that asked the same American adults about participation in different forms of religious activity in 2019 and six weeks into the COVID-19 pandemic. We, therefore, examine not only whether such practices as attending in-person or online religious services increased during the initial surge of coronavirus cases in the United States but also whether any such change is due to people participating at the same level at both time points or the movement of different people with varying levels of religious engagement.

Data, Measures, and Methods

We analyze data from the “Religious Practice in the Time of Coronavirus” module that was part of the National Opinion Research Center’s (NORC’s) AmeriSpeak Omnibus survey fielded between April 30 and May 4, 2020, among 1,002 American adults.² AmeriSpeak Omnibus is a representative sample of adults living in U.S. households (for more information, see National Opinion Research Center 2020). Four religious practices for which the same respondents were asked about their participation in 2019 and at the time of the survey are the focus of this chapter: (1) attending in-person religious services, (2) viewing online religious services, (3) gathering in person informally for religious activity, (4) and praying privately. We leverage this research design to analyze the data from two distinct perspectives. First, we reshape the original data, stacking respondents’ 2019 and time-of-the-survey measures to create a panel-level structure. By doing so, our initial analyses examine wave-level differences in these religious practices. Second, the original data allow us to investigate individual-level differences in these practices between 2019 and six weeks into the COVID-19 pandemic and thus explicate the composition of any observed wave-level differences. For both levels, we also investigate religious tradition differences. All analyses that follow apply survey weights to correct for known demographic discrepancies between our sample and the target population; wave-level analyses also incorporate a panel correction for repeated observations of the same respondents.

Results

We start this section by discussing wave-level and individual-level differences, respectively, including how religious tradition shapes them. Our attention then shifts to the effects of state-level religious regulations and respondent-level congregational closures on in-person religious service attendance during the initial COVID-19 pandemic. Finally, we consider whether the sharp decline of in-person religious service attendance we observe explains the parallel growth in online religious services, as a number of public commentators suggest (e.g., see Gjelten 2020; Yurieff and O’Brien 2020).

Wave-Level Findings

Overall, Figure 15.1 shows a considerable drop in both in-person religious services and informal religious gatherings, an uptick in weekly online religious services, and general stability in prayer. For attending in-person religious services (left-most panel), weekly rates fell from 21 percent in 2019 to 5 percent six weeks into the COVID-19 pandemic. Both monthly and less-than-monthly percentages for this religious activity also declined over time. These declines reflect an over 30 percent increase in those never attending in-person religious services between waves. While less stark, the same basic pattern holds for informal religious gatherings (third panel from the left).

By contrast, we see an 11-percentage-point rise in weekly viewing of streaming religious services six weeks into the coronavirus outbreak. At the same time, less-than-monthly and monthly rates of virtual religious services dropped some between waves. As for private prayer, the right-most panel shows very little change and no overall significant wave-level difference. At 47 percent, the same number of American adults engaged in this religious practice weekly both before and during the initial phase of the COVID-19 pandemic. Moreover, the remaining participation levels for private prayer do not change by more than 2 percentage points.

Figure 15.1 Wave-Level Differences in Religious Practices between 2019 and Six Weeks into the COVID-19 Pandemic. Note: Overall differences for all religious practices except private prayer are statistically significant at the p < 0.001 level based on a chi-square test; percentages do not always add up to 100 because of rounding. (Source: NORC AmeriSpeak Omnibus Survey, April–May 2020.)

The relative stability in private prayer is arguably the most surprising wave-level finding. Numerous studies suggest that people often turn to prayer or pray more often during times of crises (e.g., see Pargament 1997). Furthermore, at least one recent empirical study analyzing aggregate data of daily Google prayer searches for ninety-five countries argues that the COVID-19 pandemic induced more people around the world to pray (Bentzen 2020). This claim is based on the premise that online behavior often reflects what we do off-line, with the study finding a rise in online prayer searches (relative to all other Google searches) for most countries. Our analysis of wave-level data, however, calls into question how well online searches align with changes in prayer frequency, at least in response to the initial surge of coronavirus cases among U.S. adults.

We also analyzed the extent to which religious tradition affected wave-level differences in the four religious practices.³ While rarely considered in analyses of the relationship between religion and traumatic events (but see Beyerlein, Nirenberg, and Zubrzycki 2021; Uecker 2008), a great deal of prior research in political science and sociology demonstrates the explanatory power of religious tradition (e.g., see Putnam and Campbell 2010; Steensland et al. 2000). Looking at religious traditions represented in our survey data—Black, evangelical, and mainline Protestant, as well as Catholic and the nonreligious—at the wave level reveals three key patterns (results available upon request from the authors). First, we observe religious tradition variation in each of the four practices before the COVID-19 pandemic. Namely, evangelical and Black Protestants tended to be more active than were Catholics and mainline Protestants, consistent with other research (e.g., see Shelton and Emerson 2012). This variation persisted six weeks into the coronavirus outbreak. Second, despite different starting and ending points in levels of religious activity for the different traditions, the same overall pattern of change held for all of them. That is, each religious tradition experienced a steep decline in attending religious services in person and gathering informally together for religious activity, especially weekly participation, a rise in weekly viewing of online services, and stability in private prayer. Third, the magnitude of these changes varied among religious traditions. For instance, evangelical Protestants experienced the largest gain in weekly streaming of online religious services (33 percent before and 52 percent during the pandemic). By comparison, we observe no more than a 15-percentage-point increase among Black Protestants, mainline Protestants, and Catholics for this religious activity.

Individual-Level Findings

The wave-level differences raise questions about movements in people’s participation levels. For instance, is the considerable uptick in those never attending in-person religious services between waves due to already infrequent attenders simply choosing not to attend at all during the initial coronavirus outbreak, or are more frequent participants suddenly opting out of in-person religious services? And to what extent do the more stable wave-level findings reflect people engaging in the same level of religious activity before and six weeks into the pandemic? To answer such questions, we turn to individual-level data on religious practices before and during COVID-19.

Four broad patterns emerge in Figure 15.2. First, the pandemic did not generally influence the behavior of those who reported nonparticipation in 2019. For example, 90 percent or more of the nonengaged in 2019 remained so six weeks into the coronavirus outbreak, with the exception of virtual religious services. Second, a substantial number of respondents who attended in-person religious services and gathered together informally for religious activity prior to the initial stage of the COVID-19 pandemic did not continue to do so after it. Over three-quarters of Americans attending in-person religious services weekly in 2019, for instance, no longer did so six weeks into the crisis.

Figure 15.2 Individual-Level Differences in Religious Practices between 2019 and Six Weeks into the COVID-19 Pandemic. Note: Overall differences for all religious practices are statistically significant at the p < 0.001 level based on a chi-square test; percentages do not always add up to 100 because of rounding. (Source: NORC AmeriSpeak Omnibus Survey, April–May 2020.)

Third, looking at the diagonals for the greater-than-never participation categories, we see much higher percentages for online religious services and private prayer. For them, more Americans stayed engaged at prepandemic levels, especially weekly. Over 90 percent of people who prayed weekly in 2019, for instance, continued to do so six weeks into the coronavirus outbreak. Last, we observe some gains in participation frequency across the four different religious practices. The biggest ones are for less-than-monthly and monthly online religious services. We see at least a 40-percentage-point climb in these frequency categories during the initial surge of U.S. coronavirus cases.

Taken together, the results in Figure 15.2 indicate that a good number of Americans maintained prepandemic levels of religious practices six weeks into the COVD-19 pandemic; this was especially the case for those never participating before and weekly participants in virtual religious services and private prayer. At the same time, we see some movement of people in both directions. In-person religious services and informal religious gatherings experienced considerable downward mobility in participation, while some upward shifts occurred for online religious services and private prayer.

Do the above patterns vary by religious tradition? In results not shown but available upon request from the authors, we find that the individual-level findings for religious traditions generally adhere to the same basic patterns found in the full sample. That said, the extent of some of the changes differs across the various religious traditions, particularly for the two in-person measures. For instance, over one-fifth of evangelical and Black Protestants who attended in-person religious services weekly in 2019 continued to do so a month and a half into the initial surge of coronavirus cases. The comparable numbers for mainline Protestants and Catholics are 6 percent and 10 percent, respectively. On the other hand, mainline Protestants, at 42 percent, had the highest weekly continuity rate for informal religious gatherings. Black Protestants ranked second, with more than one-third of members of this religious group gathering together informally for religious activity weekly both before and during the initial stage of the pandemic.

These findings both confirm and add nuance to media accounts focusing on evangelical Protestants for defying state mandates and attending in-person religious services during the early days of the coronavirus outbreak, often in the name of religious freedom (e.g., see Collier, Trevizo, and Davilla 2020). At least among prepandemic weekly attenders, Black Protestants had the same rate of attendance six weeks into the COVID-19 pandemic as did their evangelical counterparts.

State-Level Religious Regulations and Congregational Closures

Based on the advice from public health officials to stem the spread of the coronavirus, many congregations across the country closed their doors either voluntarily or in response to state-mandated lockdown orders. To investigate the impact of this reduced access to congregations on in-person religious service attendance six weeks into the COVID-19 pandemic, we collected, read, and coded executive orders for all fifty states and then merged this coding with the survey data.⁴ Our coding scheme differentiated among four broad types of regulation operating at the time of the survey.⁵ These ranged from allowing religious gatherings or services with no explicit limits, on the one hand, to their total prohibition, on the other hand. The middle categories included executive orders that allowed gathering for religious services but either encouraged or required restrictions, such as limiting the number of people who could assemble together in person.

Figure 15.3 State Religious Regulations Six Weeks into the COVID-19 Pandemic (Source: NORC AmeriSpeak Omnibus Survey, April–May 2020.)

Figure 15.3 shows the distribution of these four categories across the United States. Total prohibitions are most prominent on the coasts. Only five states allowed religious gatherings with no limits six weeks into the COVID-19 pandemic: Florida, Tennessee, Michigan, Oklahoma, and South Dakota. The other two religious regulations—allowed with limits, encouraged or required—are generally distributed evenly among the remaining states.

What effect, if any, did these state-level religious regulations have on Americans’ in-person religious service attendance rates a month and a half into the initial coronavirus outbreak? We link the data presented in Figure 15.3 with our survey respondents’ state locations to analyze the relationship among those who attended in-person religious services at some level in 2019.⁶ Interestingly, in results not shown but available upon request from the authors, we find that the four different types of state-level religious regulations were not a significant factor accounting for variation in prepandemic attenders’ rates of attending in-person religious services six weeks into the COVID-19 pandemic.

What about closure of respondents’ congregations? Did that make a difference in whether U.S. adults who previously attended in-person religious services continued to do so during the initial stage of the COVID-19 crisis? Figure 15.4 demonstrates that it clearly did. Among prior attenders who lost access to their congregations, over 80 percent never attended an in-person religious service six weeks into the coronavirus crisis, while only 6 percent did so weekly (roughly the same number attended less than monthly and monthly). We observe a different story for prepandemic attenders whose congregations stayed open. More than three-quarters of them gathered for in-person religious services at some level during the initial surge of coronavirus cases. Over 40 percent of 2019 attenders whose congregations remained opened attended in-person services weekly a month and a half into the first wave of the COVID-19 pandemic. Another one-quarter attended monthly, and nearly one out of every ten prior attenders gathered in person for religious services at least once during the initial outbreak of the coronavirus. With that said, more than one-fifth of Americans who attended in-person services in 2019 attended no in-person religious services at this time despite having the opportunity to do so.

Congregational closures help shed light on the previously discussed religious tradition differences in attending religious services in person. Recall that among prepandemic weekly in-person attenders, evangelical and Black Protestants (at 21 percent each) were more likely to continue attending in-person weekly services during the initial COVID-19 crisis compared to Catholics (10 percent) and especially mainline Protestants (6 percent). Across all religious traditions, we find that the majority of 2019 weekly attenders’ congregations, at 85 percent, closed in response to the first wave of coronavirus cases. That said, prepandemic weekly in-person attenders connected to evangelical and Black Protestant communities had greater access relative to others. About one-fifth and one-third of evangelical and Black Protestants who attended in-person religious services weekly in 2019, respectively, reported that their church doors remained open six weeks into the COVID-19 pandemic (results available upon request from the authors). By comparison, the opportunity to attend in-person religious services was considerably lower for Catholic and mainline Protestant prepandemic weekly in-person attenders, at 7 percent and 4 percent, respectively.

Shifts from In-Person to Online Religious Services

Finally, we turn to the question of whether the rise of virtual religious services in the time of the coronavirus crisis reflects a substantive change to the U.S. religious landscape. Journalists and other public commentators speculated early on that streaming religious services were attracting new viewers, but were these largely those who shifted from in-person to online attendance, or did virtual services also draw in the previously nonengaged (M. Clark 2020; Sherwood 2020)?

Figure 15.4 Relationship between Closure of Individuals’ Congregation and In-Person Religious Service Attendance Six Weeks into the COVID-19 Pandemic. Note: Overall differences are statistically significant at the p < 0.001 level based on a chi-square test; results for 2019 attenders only; percentages do not always add up to 100 because of rounding. (Source: NORC AmeriSpeak Omnibus Survey, April–May 2020.)

Figure 15.5 Relationship between Types of In-Person Religious Service Attenders and Increased Virtual Viewing of Religious Services Six Weeks into the COVID-19 Pandemic. Note: Overall differences are statistically significant at the p < 0.01 level based on a chi-square test; results for those reporting an increase in online religious services between 2019 and six weeks into the COVID-19 pandemic; percentages do not always add up to 100 because of rounding. (Source: NORC AmeriSpeak Omnibus Survey, April–May 2020.)

As Figure 15.5 shows, nearly 80 percent of the never-to-greater-than-never and less-than-monthly-to-monthly-or-more upward shifts in online services are concentrated among those who attended in-person religious services before but not during the initial stage of the COVID-19 pandemic. At 61 percent, the attender-to-nonattender in-person group accounts for over half of the monthly-to-weekly positive change in virtual viewing. It is also noteworthy that American adults who did not attend in-person services both before and during the first outbreak of coronavirus cases compose 15 percent of the never-to-greater-than-never upward shift in online services.

Conclusion

This chapter presented some of the first survey data comparing rates of various religious behaviors before and after the initial surge of coronavirus cases in the United States (see also Chapter 16). Our analyses confirm and challenge conventional thinking about how the COVID-19 pandemic affected Americans’ religious practices in the short term. As anticipated, we observed a substantial decline in attending religious services in person and a general rise in the viewing of online religious services during the early stages of the pandemic. Somewhat surprisingly, however, the frequency of private prayer remained stable over time.

Our findings also underscore that individuals were actively navigating the shifting nature of the religious terrain and making choices about how, if at all, to practice their faith during and in response to the crisis. Broad state restrictions were not a significant variable explaining actions related to in-person religious activity.⁷ The closure (or not) of congregations to which Americans were personally connected had a strong effect, however. People were much less likely to continue to attend in-person religious services—especially at high levels—six weeks into the initial coronavirus outbreak when their faith communities closed for public safety. Furthermore, highly active evangelical and Black Protestants in 2019 remained so during the early stages of the pandemic much more than did those from other religious groups, in part, because their congregations were more likely to remain open.

It is also clear that this factor alone does not fully explain the variation we observe. Nearly one in five Americans actively sought out another place to attend in-person religious services when their regular congregation closed. And a similar number of those whose faith communities remained opened chose not to attend in-person religious services. These findings motivate future study of the determinants of religious behaviors during and following large-scale disasters that fundamentally limit in-person access to faith leaders and organizations.

What other factors are at work, influencing whether—and if so how—people practice their faith during a public health crisis? For instance, do people’s anxiety about the pandemic or direct experience with the virus influence how willing they are to look for other religious spaces holding in-person religious services when their congregation closes? Are face-to-face religious interactions so important during a pandemic that they even cross denominational lines to secure them?

We also need more data to evaluate the durability of changes. Our survey data focused on religious responses during the initial month and a half of the COVID-19 pandemic. Future studies should explore whether the patterns we observe persist. As the positive case count and death toll rose in the United States before vaccination, did rates of online religious service attendance climb further? Perhaps frequency of private prayer surged months into the pandemic as uncertainty and loneliness became the norm. Did more Americans attend in-person religious services as state regulations loosened and ended at different times and in different ways across the country? And will we witness a lasting transformation of the U.S. religious landscape because of the COVID-19 pandemic?

In particular, journalists and pundits speculate that previous in-person attenders who switched to online services will continue to attend virtually long after the pandemic subsides. Nevertheless, the above-mentioned Pew Research Center poll from July 2020 suggests such speculation may be premature; more than half of U.S. adults reported they plan to resume going to religious services in person as often or more often as they did before the outbreak once it is safe to do so again (Pew Research Center 2020a). As more and more congregations are opening their doors again but encouraging masking and other policies to protect and save lives, will people maintain this commitment to return to in-person services?

To conclude, our chapter captures important short-term consequences of the COVID-19 pandemic for religious practices in the United States. As such, it provides a useful starting point on which future studies analyzing changes in religious behaviors can build, especially those with data spanning longer periods. What short-term shifts that we observed endure over time? Much work remains to be done about how the particular—and unprecedented in the modern era—dynamics of the COVID-19 pandemic are shaping how people practice their faith and what this means for the future of religion, not just in the United States but across the globe.