When Alexis de Tocqueville penned his seminal Democracy in America lauding the strength of American communities, institutions, and associations, he noted the unique status of religion. Religion, he wrote, “must be regarded as the foremost of the political institutions of that country; for if it does not impart a taste for freedom, it facilitates the use of free institutions.”[1] Over a century and a half later, reviewing the state of American community life, Robert Putnam affirmed Tocqueville’s observation. In his magisterial Bowling Alone: The Collapse and Revival of American Community, Putnam noted that “faith communities in which people worship together are arguably the single most important repository of social capital in America.”[2]

Scores of studies have affirmed the importance of religion for cultivating social capital. This research has documented relationships between religious participation and a host of prosocial activities and positive outcomes, such as civic engagement,[3] family formation,[4] volunteering,[5] and better health.[6] And although some commentators are quick to point out that religion can be a source of exclusion and divisiveness, the balance of the evidence suggests that religious institutions strengthen the social fabric of local communities—offering community members social support, avenues for civic engagement, and often-vital social services.

Much of this research, however, treats religion as monolithic, obscuring important variation between religious traditions and denominations. Comparisons between believers and nonbelievers or churchgoers and nonchurchgoers, though often significant, fail to account for substantial cultural, theological, institutional, and socioeconomic diversity across different groups of religious adherents.

Some research has investigated the relationship between the prevalence of certain religious traditions—such as mainline and evangelical Protestantism—and specific indicators of social capital, such as crime rates,[7] civic engagement,[8] or family life.[9] But to our knowledge, no research has investigated the relationship between the prevalence of particular religious groups and overall levels of social capital.[10]

Using the Social Capital Index developed by the Joint Economic Committee (JEC) of the US Congress and state- and county-level data from the 2020 US Religion Census (USRC), we find that adherents of certain religious traditions and denominations are overrepresented in places with high or low levels of social capital. Specifically, we find that areas with high shares of mainline Protestants and members of the Church of Jesus Christ of Latter-day Saints (LDS) tend to have the highest rates of social capital, whereas areas with high shares of evangelical and black Protestants have the lowest social capital. In this paper, we offer an initial exploration of the extent to which these cross-geography associations may or may not reflect a causal relationship between participating in a given religious tradition and enjoying more valuable social capital. In future analyses, we will consider other findings reported below. We provide our data for other researchers interested in analyzing related questions.

Which Traditions Are More Common in High– and Low–Social Capital Places?

Establishing the basic associations between adherence to different religious traditions and social capital levels has been impeded by a lack of data. The JEC and USRC data assembled in the past five years allow us to examine these relationships.

JEC Social Capital Index. The JEC’s Social Capital Index is one of few comprehensive measures that captures the strength of social well-being and the vitality of associational life in America. It includes both a state- and county-level index, each of which consists of various subindexes measuring distinct facets of social capital.[11]

The first two subindexes relate to the most foundational unit of social order: the family. As Yuval Levin argues, stable and loving family attachments are the nexus of social capital development—modeling attachment to a variety of other social and cultural institutions and inculcating prosocial norms at the most foundational level of our development.[12] Because both the stability and the quality of family life are integral to our development, the Social Capital Index includes measures of both family unity and family interaction. Whereas family unity accounts for the strength and stability of family formation, family interaction measures the amount of quality time parents spend with their children.

In addition to family measures, the Social Capital Index accounts for the strength of individuals’ relationships outside the household and their connection with the broader community. The social support subindex measures the strength of Americans’ intimate personal relationships, such as the quantity and quality of close friendships. It also includes generalized social trust, or the degree to which we think our neighbors and fellow Americans can be trusted.

Moreover, the community health subindex captures the extent to which Americans are active participants in their communities through volunteerism, local civic engagement, or cooperation with neighbors. The strength of our social capital is not predicated exclusively on our most intimate social ties among friends and family; it also depends on the degree to which we trust and engage with the broader community.

The final three subindexes concern not only the strength of larger yet less tangible social institutions that are important for tackling large social problems collectively but also the willingness and ability of communities to work together to help each other and maintain order. The institutional health subindex measures civic engagement in national politics and trust in institutions such as the media, corporations, and public schools. The philanthropic health subindex measures the share of respondents who donate to charity, and the collective efficacy subindex accounts for the prevalence of violent crime.

Each of these seven subindexes captures distinct but similarly important dimensions of social capital. Places characterized by strong families, warm friendships, high trust, generosity, healthy institutions, and collective efficacy offer untold benefits—lowering the transaction costs of life and providing needed social support. Tellingly, places that score high or low on one subindex tend to also score high or low on the others.

2020 USRC. To measure religious adherence rates, we rely on data from the USRC, a census of nearly all religious institutions in America, reported down to the denominational and county level. The USRC includes data on 372 denominations, classified into 10 different traditions.[13] To stay consistent with the sociological literature, we condensed these traditions into seven mutually exclusive groups: mainline Protestant, evangelical Protestant, black Protestant, Catholic, none, Jewish, and other.[14] The USRC defines “adherents” as “all members, including full members, their children, and the estimated number of other regular participants who are not considered  communicant, confirmed, or full members.”[15]

Our results in Tables 1 and 2 show the correlations between the share of a state’s population adhering to different religious traditions or denominations and that state’s social capital index. Table 1 includes simple bivariate correlations across states, and Table 2 includes correlations at the county level. (We include population-weighted correlations in our data file.) The top portion of each table displays correlations for religious traditions, and the bottom portion displays the results for the 20 most common religious denominations, ranked by their total number of adherents.[16]

In examining these results, certain relationships run in stark opposition to conventional wisdom. Although research finds that individual religious participation is positively correlated with various dimensions of social capital,[17] states with high shares of religious adherents tend to have lower levels of social capital. Conversely, states with higher shares of non-adherents, or those who do not subscribe to a particular religious tradition, tend to have the highest levels of social capital. Importantly, however, these correlations lose all their strength when measured on the county level, suggesting that religious adherence itself may not be driving the relatively low levels of social capital in various (mostly southern) states. (However, JEC found that its county-level index is less strongly associated with a range of variables than is the state-level index.)

Earlier research by the JEC Social Capital Project found that other religious indicators were only weakly related to social capital.[18] These indicators included religious adherence rates as of 2010, congregations per capita, and participation in a religious group. Frequency of church attendance was negatively correlated with social capital across states.

Why might religiosity be only weakly associated with social capital at the county or state level? JEC hypothesized six possible explanations:

First, it may be that people who live in communities that have low social capital and substantial social and economic problems are drawn to religion as a source of support and a way of making sense of the world. Second, it may be that communities that are low in social capital for historical or demographic reasons may also be more religious for historical or demographic reasons. For instance, the institution of slavery may have had a lasting effect on the social capital levels and religiosity of African Americans. Third, secular social capital may “crowd out” religious social capital, so that places with robust community life that does not revolve around religious institutions find religious participation less valuable. Fourth, it may be that religious adherents are withdrawing from the broader civil society as it becomes more secularized (though it seems less likely that they would withdraw from their own families). Fifth, it may be that within a state or county, places with high religiosity and with low social capital are largely separate. That separateness might get obscured in aggregating up to the state or county level. Sixth, perhaps in some low-social-capital communities there is high religiosity within faith traditions that emphasize a personal relationship with God rather [than integrating faith with community] and that do not produce institutions of civil society to as great an extent.[19]

This last explanation raises the possibility that rather than religious adherence always being beneficial for social capital, the relationship depends on specific features of religious traditions. Our results confirm that, despite the ambiguous relationship between overall religious adherence and social capital, geographies with high shares of certain types of religious adherents are much more likely to be rich in social capital at both the state and county level.

Specifically, states with high shares of mainline Protestants and “other” religious groups have substantially higher social capital than do states devoid of such adherents. High levels of social capital in states with many mainliners appear to be driven by especially high levels of social support, community health, and institutional health, suggesting that these states are characterized by high levels of both interpersonal and institutional trust. Moreover, the strong positive correlation between the share of “other” adherents and level of social capital is almost entirely driven by Utah. The overwhelming majority of Utah’s population are LDS, which is classified as an “other” religious tradition. If Utah is excluded, the correlation between the adherence rate of “others” and state-level social capital falls to nearly zero.[20]

Importantly, these relationships are also observable—and are sometimes even stronger—at the county level. That is, counties with higher shares of LDS adherents and mainline Protestants also tend to have the highest levels of social capital, suggesting that there may be something unique about these religious traditions that encourages social capital development in communities.

Unlike regions with high shares of prosocial religious adherents, areas with relatively high shares of black and evangelical Protestants tend to have particularly low social capital. Indeed, states with high shares of both traditions exhibit low levels of social capital across almost all subindexes. That these correlations are similarly strong at both the state and county level suggests that certain religious traditions may fail to create and spread social capital to the same extent as other traditions.

Across almost all distinct aspects of social capital—community health, family unity, and social support, for example—areas with high shares of mainline Protestants appear to be best off, and areas with high shares of evangelicals worse. However, these correlations are just that—correlations—and are not convincing evidence that being a member of a particular religious tradition significantly influences one’s social behavior.

The remainder of this paper considers the main social capital disparities we have highlighted, using them to illustrate potential explanations for the correlations we find. We first dive into the contrasting results for mainline Protestants and evangelicals. Here we highlight the role of doctrinal differences. A robust sociological literature on religious groups has documented how varying behaviors, attitudes, and beliefs of competing traditions substantially influence the quality and quantity of their social connections. We then briefly turn to the LDS church as an example of the ways that institutional form can reinforce doctrine to build social capital. Before closing, we discuss how low social capital levels associated with black Protestants may have sources outside of religious tradition per se, rooted in broader societal inequalities.

Public and Private Protestantism: The Role of Doctrinal Differences

In 1970, Martin Marty identified two divergent factions of Protestantism in America. On one hand, he noted the growing ranks of what he called “private Protestants,” or those who emphasized “individual salvation out of the world, personal moral life congruent with the ideals of the saved, and fulfillment or its absence in the rewards or punishments in a life to come.” Opposite these private Protestants were the “public Protestants,” who were chiefly concerned with “the social order and the social destinies of men.”[21]

Marty classified evangelicals as private Protestants—mostly concerned about individual piety and personal salvation—while classifying mainliners as public Protestants. Paradoxically, this came at a time in which evangelicals were gaining prominence in the public square, appearing to be anything but private about their beliefs. But as Robert Wuthnow, a renowned sociologist of religion, succinctly noted, “Whereas mainline churches participated in progressive social betterment programs during the first half of the twentieth century, evangelical churches focused more on individual piety.”[22] Mainliners’ concern with social progress motivated them to be active participants in their communities, while evangelicals’ concern with individual salvation often caused them to turn inward, focusing on their personal spiritual formation or the vitality of their own congregations.

It is self-evident how these theological and cultural differences may help explain the significant disparities in social capital between mainliners and evangelicals. If mainliners are generally public-oriented joiners, and evangelicals relatively private and insular, then it should come as no surprise that the states and counties with relatively high shares of mainliners tend to have much higher levels of social capital.

These categorizations, however informed they may be, are based on little more than Marty’s observations. And although our results are consistent with Marty’s characterization, they are based on the state and county level, not on individual behaviors. In the following sections, we draw on a wide range of work and individual-level data to determine whether public and private Protestants actually differ in their attitudes and behaviors—all of which have clear implications for the development of social capital.

Community Health. Mainline Protestants earned their name not because they constituted a majority of religious Americans but because of their centrality in American culture.[23] Even though mainliners have not been a majority (or even a plurality) of religious Americans for decades and have experienced the most precipitous declines in membership, they have retained their designation because of their continued importance in the development and maintenance of core social institutions. Many of America’s most vital social institutions, such as our universities, libraries, and hospitals, remain associated with mainline Protestantism.[24] This cultural prominence led Nancy Ammerman, a sociologist of religion documenting the religious experiences of mainliners, to contend that “the terms mainline and civic have long been seen as nearly synonymous.”[25]

Sociologists and scholars of religion later quantified this observation, examining the degree to which members of different religious traditions participated in community life. Specifically, researchers found that mainline churchgoers were substantially more likely than churchgoers of other religious traditions to be members of voluntary organizations and engage in a host of other communal activities. Evangelicals, by comparison, were among the least civically engaged groups.[26]

Evangelicals tend to have different demographic characteristics than mainliners do, skewing younger, less educated, and lower income.[27] But even holding socioeconomic status constant, mainliners were about 20 percent more likely than Catholics to be a member of a nonreligious volunteer association, and evangelicals were 30 percent less likely to be a member of such group than Catholics were.[28]

Evangelicals are by no means unsociable and withdrawn from civil society altogether. For one, evangelicals are a bit more likely than mainliners to regularly attend religious services, helping them build social capital within their congregational communities.[29] Moreover, they are no less likely than mainliners to vote in national elections and are often more concerned about public affairs.[30] Political engagement, however, may be tangential to strong and healthy relationships if it is expressed mainly at the dinner table and in the solitude of a voting booth. And when it comes to allocating their social and civic energies, evangelicals are much more likely to dispense them within the walls of their own church, supporting their own congregation.

The nationally representative General Social Survey (GSS) asked respondents in 2004 about their membership in church-affiliated groups and asked the 2012 respondents whether they had volunteered in the past year. Though differences could simply reflect the different years in which the questions were asked, mainline Protestants were much more likely to have volunteered (52 percent) than to have been a member of a church group (34 percent). Evangelical Protestants were less likely than mainline Protestants to have volunteered (46 percent) but more likely to have been a member of a church group (40 percent).

Both the findings of others and these data support Marty’s hypothesis that mainliners are chiefly oriented toward public involvement and community engagement, while evangelicals tend to be more interested in engaging with their smaller, tightly knit church communities.

Family Unity. Despite clear and pronounced differences between mainliners and evangelicals regarding community participation, the distinction between public and private Protestantism has much more ambiguous implications when it comes to family life. For one, evangelical Protestants have been among the most vocal supporters and ardent defenders of traditional family values. Conversely, many mainline denominations have been much more open to the sexual and familial developments of modernity, suggesting that evangelicals may be champions of the family, while mainliners focus on community. 

Therefore, it may be quite surprising to see that areas with high shares of evangelical Protestants tend to have the worst family outcomes and areas with many mainliners the best. Sociologists have observed these trends in prior work, including our colleague, W. Bradford Wilcox. Wilcox argues that mainline adherents tend to “talk left and walk right,” whereas evangelicals tend to “talk right and walk left.”[31] In other words, when it comes to the family, evangelical and mainline Protestants demonstrate some cognitive dissonance with respect to their purported views and actual practices.

No group of Americans has mounted a stronger opposition to the development of modern familial and sexual ethics than evangelicals. More than any other religious tradition, evangelical Protestants oppose the dissolution of traditional gender roles, the rise of divorce, the growth of out-of-wedlock child-rearing, and the increase in premarital sex.[32] Mainliners, on the other hand, have been comparatively more open to these developments—evidenced by their inclusion of women in their clergy, recognition of same-sex marriages, and destigmatization of single parenthood.[33] But when it comes to their actual marital and sexual practices, mainliners appear far more conservative, and evangelicals appear far more progressive (Figures 1 and 2).

The trends in values and behavior are in the same direction for both mainline and evangelical Protestants, with both becoming more permissive and having more family instability. But across the nearly 50 years, evangelicals are substantially more likely to embrace traditional family values in theory (Figure 1), even as they are slightly more likely not to actually carry out those ethics in their relationships (Figure 2). Specifically, a plurality of evangelicals maintain that premarital sex is always wrong, whereas only 13 percent of mainliners agreed in the most recent year of data. But evangelicals have actually been more likely than mainliners to divorce, even though mainliners are on average significantly older.

Although premarital sex and divorce are two distinct issues in family ethics and therefore might not expose the tension between belief and practices convincingly, other research has documented a similar divergence on several other questions related to family values. Evangelicals, for example, are more likely to cohabit before marriage and have sex at earlier ages compared to mainliners, despite being more opposed to those practices in theory.[34]

Put succinctly by Wilcox, evangelicals “have largely succeeded in articulating and fostering a distinctively familistic ideology among their members. [But] the conservative Protestant record of success when it comes to shaping behavior is not uniformly good.”[35]

Granted, mainliners tend to be more educated and have higher incomes than evangelicals do, and both characteristics are associated with embodying and exhibiting bourgeois values. Nevertheless, the causes may not be merely socioeconomic. The relative dearth of community connection among evangelicals may weaken the moral guardrails that help translate shared values into congruent individual behavior. If so, it could explain why areas with many evangelicals tend to have lower levels of family unity.

Social Support. The strength and quality of close social relationships are the bedrock of social capital. One virtue of private Protestantism is its ability to forge close personal ties and intimate relationships, an invaluable element of social capital development. Conversely, public Protestants’ vigorous engagement with the broader community may come at the expense of cultivating intimate relationships.

Even though state-level correlations show that areas with many mainliners offer more robust social support, while the reverse is true of evangelical prevalence, individual-level data suggest that evangelicals are just as social as mainliners. Specifically, both groups of Protestants are equally likely to spend a social evening with a friend, relative, or neighbor more than once per month (Figure 3).

Where mainliners significantly outperform evangelicals, however, is in generalized social trust. Generalized social trust measures individuals’ trust in a generalized “other,” or their neighbors, community members, or fellow Americans. Generalized trust, unlike trust in close friends or relatives, is a particularly strong predictor of a host of positive personal and communal outcomes, such as self-rated health, happiness, and volunteering.[36] And when it comes to this sort of trust, evangelicals are substantially more wary than mainliners are (Figure 4).

To be sure, social trust is dependent on the general trustworthiness of those in the community. If evangelicals are more likely to interact with people who are less trustworthy, then their lower levels of trust might be perfectly justified. And given that evangelicals are much more likely to live in high-crime areas compared to mainliners, their deep-seated distrust seems at least partly justified.[37] However, it is not clear which way the causal arrows point—whether low social trust inflates crime rates or whether crime diminishes social trust. Chicken-and-egg questions notwithstanding, evangelicals are generally less trusting in others, which can significantly hinder social capital development beyond kin and close friends.

Moreover, recent research has affirmed the fact that areas with high concentrations of evangelicals tend to have much stronger in-group bonds but generally weaker out-group bonds. In a recent paper, Opportunity Insights leveraged billions of social media connections to find that areas with high shares of evangelicals tend to have the highest levels of “cohesiveness,” meaning that they have dense, tightly bound social networks. But such areas also have low levels of civic engagement and are less likely to host many friendships that cross socioeconomic boundaries.[38] Areas rich with mainline adherents exhibit inverse tendencies, further supporting the public and private Protestant hypothesis.

The balance of the evidence suggests that, although evangelicals’ personal social lives are just as robust as mainliners’, they are generally less trusting in others. And significantly lower levels of trust greatly inhibit evangelicals’ abilities to build stronger ties in the wider community.

LDS: Doctrine Plus Institutional Design

Doctrine plays an important role in explaining the high social capital levels and community-mindedness of LDS members. For instance, a JEC report examining Utah’s off-the-charts social capital notes that LDS doctrine teaches that “marriage and family relationships are eternal and thus should be built to endure” and cites research that premarital sex is much less common and marriage more common among LDS members than among members of other religious faiths.[39]

The report attributes high levels of family interaction, in part, to the fact that “members [are] encouraged to set aside one night a week for ‘Family Home Evening.’”[40] Community-mindedness is promoted via other doctrinal features. Members are expected to tithe, contributing a tenth of their earnings to the church. Further, the JEC report explains that “Latter-day Saints are also encouraged to help the poor through fasting once a month and then contributing at least the amount they would have spent on the forgone meals to offerings that fund the Church’s welfare programs.”[41] The church also strongly encourages volunteerism.[42]

Other features of the LDS church, however, are institutional matters. One reason social support among LDS church members is high has to do with the way the church formally structures those relationships. Per the JEC report:

Latter-day Saint congregations (called “wards”) are structured geographically, with those living closest to each other generally being members of the same congregation. Thus, Utahns are often not only neighbors with those living around them, but they frequently are fellow congregants with many of their neighbors as well. . . . Besides the overlapping nature of neighborhoods with wards, the Church also promotes service among ward members by assigning each member of the ward to look after specific individuals or families within their ward.[43]

This structure surely contributes to the finding that over seven in 10 LDS members “feel closely connected to their neighborhood and the people who live there,” according to AEI’s Daniel A. Cox. Among Americans as a whole, only half feel that closeness. LDS members are far more likely than other Americans to socialize with friends and neighbors in their homes.[44]

Community health among LDS members is facilitated by an infrastructure of social welfare provision, including grocery stores for the poor, thrift stores, social services, and job training programs.

Social capital investment infuses the LDS church to an extent matched by few other religious traditions in America.

Black Protestants: The Importance of Broader Societal Inequities

Had white supremacy played a minor role in American history, “black Protestant” would not be a meaningful category in our analyses. Black Protestants resemble white evangelicals in their religiosity and fundamentalism.[45] But the reality of southern segregation meant that black evangelicals created their own religious institutions that developed differently from those of white evangelicals. The church has been much more a center of community life among black evangelicals than for their white counterparts, serving as a refuge from an often-hostile wider world and a center of organizing during historic civil rights struggles.

Black Protestants, therefore, tend to be more community-oriented than white evangelicals. All else equal, then, we would expect that places where black Protestants are prevalent would have higher social capital than places where evangelical Protestants are prevalent. (Note that the traditions labeled “black Protestant” and “evangelical” can and do include both white and black Americans.)

However, separate is rarely equal. On the one hand, the GSS data confirm that black Protestants are more likely than evangelicals and mainline Protestants to belong to a church group (50 percent, compared with 40 percent for evangelicals and 34 percent for mainliners). They are also more likely than members of either group to spend a social evening with a family member, friend, or neighbor more than once per month (68, 44, and 34 percent, respectively). On the other hand, social trust is strikingly low among black Protestants. Just 12 percent indicated in the 2018 GSS that people in general can be trusted, versus 31 percent of evangelicals and 39 percent of mainline Protestants.

The relatively low social capital associated with black Protestants is almost surely more related to the societal inequities that have impeded and shaped African Americans over the past 400 years. Across states and counties, the JEC Social Capital Project found correlations between the share of black residents and indicators of low social capital that are similar to the correlations we find here for black Protestants.[46]  If anything, the correlations are stronger (more negative) when looking only at African American shares rather than black Protestant shares (especially at the county level), suggesting that black Protestants have higher social capital than blacks do generally.

That historic racial inequality is behind these social capital disparities is suggested by Figure 5, produced by the JEC Social Capital Project, which overlays family unity subindex scores in the southeastern US with data on slave prevalence in 1860.[47]

Figure 5. Social Capital Index and Historical Prevalence of Slaveholding

Source: Social Capital Project, “Social Capital, Slavery, and the Long Reach of History,” Joint Economic Committee, https://www.jec.senate.gov/public/index.cfm/republicans/2019/1/social-capital-slavery-and-the-long-reach-of-history.

Conclusion

For too long, policy analysis has neglected the relationships between various outcomes and social capital and between social capital and various social influences. So, too, has policy analysis tended to focus on economic outcomes and economic explanations. The roles of culture and institutions have been downplayed, in part due to measurement challenges.

The relative infancy of social capital measurement and research leads us to make inferences from simple correlations in these analyses. The extent to which correlation may be taken to reflect causation is, of course, open to debate and requires deeper analysis. We hope that by providing these initial analyses and our data we can inspire further research leveraging more sophisticated identification strategies. Understanding social capital and the influences of culture and institutions is crucial for expanding opportunity in America. We will return to this theme in future analyses.