Although political science once ignored the role of religion on politics, today the heightened attention given the topic is demonstrated by a number of book series by academic presses. Both of these two new volumes in the Georgetown University Press religion and politics series use survey research to explore the impact of various aspects of religion on the attitudes and behaviors of ordinary citizens.

Gregory Allen Smith's Politics in the Parish explores the impact of Catholic priests on their parishioners. The book is an unusual read, starting with complex statistical models, then turning to interview material from priests in Virginia, Washington, DC, and Maryland, and then concluding with complex statistical models. Each section is strong, and taken together they constitute an important contribution to an understanding of the role of priests in political life.

I was delighted to see Smith begin with a sophisticated secondary analysis of the Notre Dame Study of Catholic Parish Life, a rich survey from the 1980s that has been underexplored. Using hierarchical linear models to combine data from priests and their flocks, the author begins with simple models that seek to test the impact of liberal and conservative priests on the ideology and partisanship of their parishioners, and then on a variety of concrete issues addressed by the church that go beyond abortion and allowing gay and lesbian teachers in classrooms—including poverty, racism, justice, and peace.

What is impressive in this analysis is how Smith builds on the simple models, considering interactions with assessments of the priest's homily quality and members' willingness to accept guidance on these issues. He concludes that liberal priests have significant influence on many issues, but that conservative priests have little or even a negative impact on attitudes.

The middle section of the book includes some fascinating interview material from individual priests and content analysis of homilies and church bulletins. Here, Smith creates a portrait concerning how often each priest discusses particular issues and how they are discussed. He then links this information to surveys of congregations, again using hierarchical models. Smith tests for a variety of interactions but concludes that in 2004 priests appeared to have no direct impact on the voting decisions or political attitudes of their parishioners. But they do shape their flock's willingness to accept religious guidance on political matters and their particularistic dedication to the Catholic Church, and these provide indirect paths of influence. I was somewhat surprised by the consistency of the null result in this section, especially since congregants might choose a parish because they agree with the priest.

The book might have been strengthened by considering the constraints that parishioners may put on priests' political exhortations. Other scholars have shown in qualitative and quantitative work that Protestant congregations constrain their pastors. Priests are leaders in a hierarchical religious institution, and so it may be that there is less influence of congregations on priests, but priests may temper their homilies on the basis of perceived reactions as well. But overall this is a fine book, one that allows us to understand individual priests, their efforts to engage in political discussion, and the impact of these efforts and that nicely employs sophisticated hierarchical modeling.

In Pews, Prayers, and Participation, a team of established scholars seeks to understand the connection between religion and civic responsibility in the United States. The authors begin with a review of the work by political theorists and empirical scholars on the subject, and then systematically explore the topic using a truly impressive array of surveys. The result is a significant contribution, both for the evidence it provides and for the questions it provokes.

The authors define civic responsibility as “virtuous civic engagement,” which is, in turn, operationally defined as a set of behaviors, capacities, and virtues. Responsible behaviors include joining civic associations and giving time and money to charitable causes. Responsible capacities are political and civic skills, and virtues include obeying the law, feeling efficacious, and having high levels of social trust and tolerance.

It was not always clear to me how some measures were related to responsibility. If churches enhance civic and political skills, those skills could be used responsibly or irresponsibly, and arguably a religiously responsible citizen might be obliged to violate unjust laws. Moreover, some concepts are measured better than others—a question on how people get ahead in America does not really measure a work ethic. But the authors must use what data they can find, and I like the fact that they approach these questions broadly.

The central concept is the form of religious expression, which the authors typologize on the basis of measures of public and private religiosity. Those who attend church less than monthly and pray less than daily are said to have a “diminished” form of expression, while those who both attend and pray more often are categorized as having an “integrated” form of expression. Those with high attendance and low rates of prayer fall into the “public” religious form, and those with the opposite characteristics have a “privatized” religious expression. I could quibble with these measures and cut points, but the typology makes sense.

The authors then approach each dimension of civic responsibility, as well as each measure of these dimensions, first by showing differences across the categories of religious expression in bivariate analysis, and then in multivariate models that hold constant some basic demographic variables and a measure of religious tradition (denominational families). From these analyses, they conclude that religion has a generally positive impact on civic responsibility. The effects are most clearly demonstrated for contributions of time and money to charitable causes, a reasonable result since many congregations sponsor or are involved with charitable activities. The results are less impressive for civic virtues, where the data show those with integrated religious expression to be no more socially trusting and to be less tolerant than the least religious.

The bivariate tables show results from many surveys. In some cases the surveys tell a consistent story, but in other tables the surveys show different results, and these are presented as equally plausible. The authors draw conclusions by comparing percentages across categories of religious expression, but there are no pairwise tests of statistical significance. Instead, the tables contain a coefficient that tests whether the overall differences across categories is significant.

This carries over to the multivariate strategy as well, for the authors use multiple classification analysis to present a single coefficient for the impact of forms of religion and of religious traditions. But these coefficients do not tell us which groups contribute to the significant relationship. A high coefficient for religious tradition, for example, could mean that Catholics have higher scores than all other groups, or Evangelicals are low and Jews are high, or many other things. Moreover, it is possible that the multivariate controls change the ordering from the bivariate analysis, but this cannot be assessed from the evidence presented. A more straightforward approach might have been to treat categories as dummy variables in more traditional linear models. And it may well be that forms of religious expression have different impacts across religious traditions, but the models do not test for interaction effects.

The authors try heroically to sort out causality, but with cross-sectional data it is difficult to be confident about the impact of religious involvement on civic responsibility. It may be, for example, that those who have integrated religious forms are more likely to join civic organizations than those with privatized religious forms, but it may also be that less socially inclined citizens are not active in religious organizations or civic ones because of personal proclivities. That said, I am sympathetic with the way the authors attempt to deal with some intractable questions regarding causality.

Overall, this is a solid contribution that deserves the serious attention of anyone who is interested in the relationship between religion and citizenship. The authors have done a remarkable job of assembling data and carefully sorting out explanations. They make a strong case that involvement in churches is the single most common form of group membership and that this has important consequences for those who join. Just thinking about working with all of these data sets is exhausting, but the book inspired me to load up some data to test these relationships further.

Both books take religion seriously, and their authors seek to think carefully about the ways that religion might matter. As the field of religion and politics has matured, it now asks more complex theoretical questions and needs more complex models to test the implications of these theories. Both of these books do a comprehensive job of exploring their disparate research questions, and both leave enough questions for other scholars to ponder. This makes them both well worth reading.