Do the religious convictions of Americans explain the so-called “God gap,” the sorting that leads the devout and the unaffiliated to their separate homes in the Republican and Democratic parties, respectively? A great deal of religion-and-politics research, not to mention most conventional wisdom, assumes a direction of influence from religious faith to political behavior. But that narrative has come under increasing challenge in recent years, especially as political scientists and others have examined the role of political backlash in eroding rates of affiliation. Could it be that the direction of influence runs the other way – that partisanship and political environment shape religion?
Michele Margolis's From Politics to the Pews is the leading edge of this way of thinking about the intersection of religion and politics in the United States. Her analysis is not only theoretically compelling, but also path-breaking in its research design and deeply profound in its implications. The basic argument is that the religiosity of Americans reflects their partisan attachments more than the other way around. Democrats are less religious because they are Democrats; they are not Democrats because they are less religious. And the same goes for Republicans, who do not necessarily gravitate to the GOP because of their religiosity, but “update” their level of religiosity because religion is associated with the Republicans. It is a provocative position, but Margolis presents a sophisticated theory and marshals considerable evidence to make her case.
Margolis grounds her theory in group-based assumptions that focus less on theological constructs than affiliations and collective behaviors, both religious and political. As a result, her two primary measures of religion are affiliation and attendance at religious services and her key measures of politics emphasize partisanship and other dimensions of group environment. While these measures are well-trod territory for religion-and-politics scholars, Margolis's lays out an innovative path by drawing from theories about identity formation during the life cycle, where she identifies two well-established insights and combines them in a novel way. She starts with the widely accepted argument in political science that partisan identification solidifies in adolescence and early adulthood and tends to stick as we move through the life cycle. She then notes a contrasting result in the sociology of religion that the same life-cycle timing does not apply in the formation of religious identity, which tends to be much more inchoate and even resistant in formative years and does not reach a settled pattern until adults begin to make decisions about marriage and family. The formative sequencing here is crucial: Margolis theorizes that Americans’ stable partisan identities help define our in- and out-groups and therefore influence decisions about “updating” our religious associations later in life.
Chapter 4 is a first pass at the theory. The life-cycle explanation requires data that tracks change over time, which she finds in two panel studies: the Youth-Parent Socialization Panel Study, which followed a cohort from 1965 to 1997; and the National Annenberg Election Study, which surveyed the same respondents from 2007 to 2014. The lagged analyses of both studies suggest that levels of religiosity were indeed partisanship-induced, even when controlling for other factors, including regional variation (Is the phenomenon really a Southern thing?), levels of civic engagement more broadly (Are Republicans just more likely to be joiners?), and genetic influences.
The next several chapters subject the theory to a range of additional challenges. Chapter 5 probes alternative explanations for religious change. Chapter 6 investigates the role of political knowledge, because the politics-shapes-religion hypothesis “assumes that Americans are aware of the religious … differences that separate the Democratic and Republican parties” (p. 131). The chapters employ an impressive array of data sources, including both mass surveys and experiments, at both a national and congregation level. The results hold up.
Chapters 7 and 8 examine the “reach” of the theory across traditions and periods of time. Margolis focuses on a specific group—African American Protestantism—that ranks high in both religiosity and Democratic partisan identity, an apparent challenge to the theory. Margolis notes that, while African Americans and white Republicans share the religious life-cycle effect, African Americans remain “faithful partisans” who are largely impervious to the religion-based strategies of the GOP. She suggests a key reason for this resistance might be that the Democratic party “represents many of the religious values emphasized in black churches” (p. 177). Margolis does not see that interpretation as a return to the conventional religion-influences-partisanship narrative because African Americans have no need to “update” their religious perspectives with their partisan identities when both are already aligned. This conclusion was a rare instance when I wondered if Margolis was close to begging the question, and in any case it remains an open question to me whether African Americans present a strong test of generalizability across traditions when both religious and partisan identities are so deeply entangled.
Margolis then takes on the argument that her overall results are time-bound. Is the life-cycle effect limited to our recent history of intense religion-based partisan appeals? She finds counterevidence in a case study of the 1960 presidential election, a religion-saturated contest that also happened to include the third wave of an ANES panel study. She finds that both Protestant Democrats and Catholic Republicans declined significantly in church attendance from 1958 to 1960, which represents some tentative evidence of partisan-induced religious change outside our own era.
My quibbles with this book are slight. My primary response is admiration for Margolis's insightful theory, innovative design, varied use of data, and strong results. I am also excited about the prospects for future investigation. Margolis suggests several of those prospects in her final chapter, “The Religious Sort,” an evocative title in this time of widespread discussion about the nature and scope of polarization.
The book does indeed provide empirical warrant for what many of us already accept, namely, that affective polarization has colonized areas of American life we have generally imagined as “non-political,” at least in a narrowly partisan way. I kept coming back to common models of civil society as associations that “mediate” between the state and individuals or provide “seedbeds” for development of social capital and civic skills. Each of those models assumes that religion, as a sphere of human experience within civil society, has a certain autonomy that enables its mediating or formative role. Margolis's rich work pushes us to re-theorize and investigate anew these and other established ideas.
