Robert Putnam’s Bowling Alone was a surprise hit when it came out in 2000. The book, an expansion of a 1995 journal article by Putnam, put forth a relatively simple argument: Social capital, as embodied in the voluntary organizations and social practices that bind people together, has been declining dramatically in the United States over the last few decades. Putnam’s glum dispatch had an enormous impact on public dialogue. The phrase “bowling alone” achieved the kind of pop-catchphrase familiarity that few academics even dream of, and an entire industry of scholars and commentators emerged to both examine and challenge Putnam’s findings and interpretations.

Now, Putnam moves from bowling alone to praying together. American Grace, co-written with David Campbell, associate professor of political science at Notre Dame, may not quite achieve the same level of success, but it nonetheless has the potential to influence the public conversation. A comprehensive study of religion in America, it is full of new empirical data from both a large survey and intensive field work conducted by Putnam and Campbell and their colleagues, and it offers provocative insights into the state of faith in America. If Bowling Alone painted an ominous picture of the fraying American social fabric, American Grace, as the name implies, offers a more optimistic image–one nation of many religious and not-so-religious traditions, but learning to live as one. Sweeping and persuasive though their study of religion may be, their analysis of its intersection with politics is less compelling.

American Grace examines a basic question: In a nation with high levels of religious commitment, how can religious pluralism coexist with religious polarization? The authors dissect these two phenomena–and the tension, or lack thereof, between them–and conclude that, yes, levels of religious identification (self-identifying with one or another religious tradition) and religious practice (for example, attending church services, praying, and saying grace before meals) remain very high. And the religious landscape is extremely diverse, with a variety of denominations, a large number of independent churches, and a range of theology, style of ritual, and governance structures across congregations. But there is a qualification: Young people today show declines in both identification with some religion and depth of religious practice. The likely result is a slow but steady decline in both religious identification and practice over time. That said, the decline will be gradual, and from very high levels, so religion will continue to be a major force in American society for a long time to come.

A second strand of their argument is the increasing religious polarization in America. They note the growth in both the proportion of the population that identifies itself as having no religion and the proportion of church attendees who are devout evangelical Protestants. In other words, we’ve seen an expansion in what might be considered the two ends of the religiosity dimension. In this finding, Putnam and Campbell see a form of religious polarization of greatest concern to them–not between denominational groups, as with the Catholic-Protestant divide that characterized some periods of American history, but the chasm between the traditionally devout of whatever denomination (defined by them in various ways, but usually either as attending church services at least once a week and/or as saying grace before meals at least daily) and the secular or minimally religious.

Putnam and Campbell argue that despite this increased polarization, there is surprisingly little religious hostility in America. They base their conclusions on the findings from a survey they conducted that presented respondents with a “feeling thermometer” about other religious groups–questions that ask how warmly or coolly respondents feel about different groups, on a scale of 0 (cold) to 100 (very warm). The vast majority of religious groups, they report, receive scores above 50 from respondents who are not members of the religious group being asked about, which indicates they are viewed more positively than negatively. (Not surprisingly, most respondents feel very warmly toward their own religious tradition.) The most highly regarded groups, according to these data, are Jews, Catholics, and mainline Protestants. Evangelical Protestants and the non-religious are viewed less favorably by those who are not members of the groups, but both groups also have average scores above 50. Mormons, Buddhists, and Muslims score below 50, with Muslims the lowest, but still above 40.

Putnam and Campbell argue that this range of feelings is relatively benign. They note that a large majority of Americans see religion as a good influence in American society, and that this perception is backed by evidence that the devout are also more involved in charitable and civic activities and general acts of neighborliness than the less religious. A large majority also see religious diversity as a good thing. They overwhelmingly believe that people of other faiths can go to heaven. For example, 93 percent of Catholics believe that non-Catholics can go to heaven, and 89 percent of those extend entry to heaven to non-Christians. (Interestingly, according to a 2007 Pew Forum poll, only 82 percent of Catholics believe in heaven.)

But if most Americans think religious diversity is a good thing, what to make of last summer’s firestorm over the proposed Islamic center in lower Manhattan and recent polls showing decreasing approval of Muslims? American Grace was written before the controversy, which might have qualified the authors’ optimism. In an August 2010 Pew Forum poll, 30 percent of respondents reported a favorable view of Islam and 38 percent an unfavorable view, with 32 percent responding that they didn’t know. These responses are less favorable than responses to a Pew Forum poll in July 2005, which asked the same questions; to that survey, 40 percent of respondents reported a favorable view of Islam and 36 percent an unfavorable view.

Putnam and Campbell would likely argue that this controversy is reminiscent of the harsh disapproval of and even violence against Catholics less than a century ago. As Catholics became more integrated into American society, and as inter-religious marriage and friendships with Catholics became more common, disapproval of them has pretty much disappeared. So, perhaps, might attitudes toward Muslims become more favorable over time. Putnam and Campbell do not make any specific predictions on this topic. My own take is that Muslims, like others in the past, will indeed become integrated into American society, but that this integration is likely to take quite a long time. And if another terrorist attack tied to Islamic extremists occurs in America, all bets are off.

But overall, Putnam and Campbell conclude that religious polarization does indeed coexist with stable and more or less tolerant religious pluralism, and then move on to ask how that can be. The answer lies in the fluidity of the religious landscape in America, the frequency with which Americans change religions and marry partners of other faiths. They note, for example, that 35 to 40 percent of Americans have changed religions at some point in their lives, and that in half of all the marriages in the 1990s, the spouses came from different religious backgrounds. Nearly everybody, they argue, has an “Aunt Susan” or a “friend Al” of another faith, making it hard for them to believe that people like Susan or Al are evil or barred from heaven. The authors use their own families as examples:

One of us (Campbell) is a Mormon. He is the product of what was initially an interfaith marriage–as his Mormon mother married his mainline Protestant father. Eventually, his father converted to Mormonism. His mother too had been a convert years before. As a child she left Catholicism to become a Mormon, along with her parents but only some of her siblings. Consequently, a reunion on either side of the family brings together a multi-religious mix.

The family tree of your other author (Putnam) also encapsulates the religious churn that is so common in America. He and his sisters were raised as observant Methodists in the 1950s. He converted to Judaism at marriage; he and his wife raised their two children as Jews. One child married a practicing Catholic, who has since left the church and is now secular. The other child married someone with no clear religious affiliation but who subsequently converted to Judaism.…It would be hard to rouse anti-Jewish or anti-evangelical or anti-Catholic or anti-Methodist or even anti-secular fervor in this group.

This pluralism, they conclude, is “America’s grace.”

So far, so good. But there is another, less explicit theme in the book that is potentially more disturbing: the relationship between religion and politics. Putnam and Campbell did not set out to write a book on political polarization; they set out to write a book on religion, which they have done, and very well indeed. It does seem to me interesting, however, to think about a very different pair of ending chapters to the book if the authors had adopted a somewhat different perspective. The second-to-last chapter of the book, “A House Divided?,” concludes that the answer to that question, at least when it comes to religious polarization, is that it’s not so divided. The last chapter, “America’s Grace: How a Tolerant Nation Bridges Its Religious Divides,” summarizes their argument about the fluidity of American religious identification and the importance of family and friendship networks that bridge religious divides.

But suppose the question about the divided house were a question about political rather than religious polarization. That house, I opine, would look very divided indeed. Putnam and Campbell themselves document that religious polarization in contemporary America–the divide between the devout and the secular–is characterized by political divisions: a polarization of attitudes on a number of value issues, and sharp differences in party identification. The devout are much more likely to identify and vote Republican–and are much more disapproving of both abortion and same-sex marriage–than the secular and minimally devout. They also note that approval and disapproval of different political ideologies is more sharply divided than feelings about other religious traditions.

They try to find some good news in their analysis of religion and politics, arguing that only two issues–abortion and gay marriage–account for the “God gap” between Democrats and Republicans. Very religious people have long been more opposed to both abortion and homosexuality than the less religious. Starting around 1980, the two parties began clearly to differentiate themselves along this dimension, which led, not surprisingly, to the observed link between religiosity and party identification. Their good news comes from a generational analysis of the “God gap” in party identification. They find that the association between religiosity and party preference is strongest among young people of the post-Boomer generation. But they also report data showing that these same young people are less divided than the overall population on the two most divisive issues: abortion and gay marriage. They are, interestingly, much more liberal than older respondents on gay marriage but more conservative on abortion. Putnam and Campbell suggest that this might signal a narrowing of the political divide between the parties. Since the issues of abortion and gay marriage defined the polarization, if they become less salient, the polarization of the two parties may become less stark as one generation succeeds another.

But that may be wishful thinking. There are many indications of the growing strength and salience of political divisions. The parties have become more ideologically differentiated, as evidenced by the large and growing differences between voters who identify with one or the other party on a whole range of issues. As Alan I. Abramowitz documented in The Disappearing Center, voters are less likely than in the past to identify as independent or moderate, and more likely to identify strongly with one party or the other.

The job approval ratings for President Obama, as reported in Gallup polls, are more sharply divided by party than those for any other president in recent times. In August 2010, 88 percent of Democrats approved of Obama compared to only 23 percent of Republicans, a gap of 65 percentage points. At similar points in their presidencies, the partisan gap for George W. Bush was 32 percentage points, for Reagan 45 percentage points, and for Clinton 52 percentage points. Additional evidence for the growth of political polarization comes from the increasingly lopsided electoral margins within congressional districts, in both presidential and congressional elections. According to Abramowitz, the number of House seats classified as safe for one or the other party rose from 122 in 1976 to 216 in 2004. Americans are increasingly separating themselves into one-party enclaves.

This increasing polarization is not good news for American democracy. It inhibits our ability to agree on and achieve common purposes. It is increasingly difficult to get anything done politically; Americans’ opinions of their legislative bodies are at historical lows.

And that leads me to wonder whether religion exacerbates the political polarization that seems so harmful. The link that does exist is subtler than one might have expected. One of Putnam and Campbell’s more interesting findings is that the relationship between religion and politics is not due to a high level of politics in sermons and other formal religious activities in church. In only a minority of congregations is there any political activity at all. They report, for example, that only 24 percent of evangelicals say that their local church has voter registration drives or distributes voter guides; even lower percentages report hearing political content in sermons. By substantial majorities, American respondents to surveys think that churches and clergy should stay out of politics, and they pretty much do. (Interestingly, the most overt politicking goes on among African-American Protestants and Jews.)

Putnam and Campbell offer an alternative, social-capital-based explanation for the relationship between religiosity and political identification. People develop friendship networks within congregations, and these networks reinforce political views. “Religious social networks,” they write, “serve as echo chambers.”

Political polarization is almost certainly intensified by geographical sorting that increasingly reflects clustering by lifestyle and ideology as well as by income, a phenomenon explored in Bill Bishop’s The Big Sort. The sorting likely takes place among congregations as well. The fluidity in religious and congregational affiliation noted by Putnam and Campbell allows for people to sort themselves among congregations by ideology and political preferences, as well as by preferences for theology and style of ritual. Then, as Putnam and Campbell note, congregations reinforce political and ideological divisions through the friendship networks and bonding social capital that develop among congregants.

Putnam and Campbell note that feelings about political ideologies are more sharply divided than feelings about other religions. In interpreting the “feeling thermometer” results, they report, “People who describe themselves as ‘very liberal’ are cooler toward conservatives (36), and ‘very conservative’ respondents are, in turn, cooler toward liberals (38) than any religious group is toward another.” This suggests that our tolerance of Aunt Susan and our friend Al works only with regard to religious differences and not political differences. (Perhaps Aunt Susan the evangelical will go to heaven but Aunt Susan the Republican will not?) Perhaps religious differences are simply less important to people than political divisions. Or perhaps they are tied together but expressed as political rather than religious if people feel it is more acceptable to articulate political divisions.

Congregational leaders are almost certainly complicit in the sorting of Americans by ideology and politics. They have learned how to market themselves in a competitive religious environment and how to cater to the attitudes and preferences of their congregations. They may not do this consciously, of course, but it would be surprising if they did not subtly tailor their messages and interactions to reflect their congregations, and thus to reinforce the dominant ideology and politics. They do little, it seems, to construct opportunities for the bridging of partisan and ideological divides.

If our congregations are indeed abetting our political divisions, then it is hard to be as sanguine as Putnam and Campbell about the “grace” of religious differentiation and pluralism. And even the religious tolerance that Putnam and Campbell tout is still a work in progress, as American Muslims can tell you. Our religious pluralism may be America’s grace, but the intensifying political polarization in American life, reinforced by our religious communities, may well be America’s sin.