Structural explanations of residential segregation usually focus on the role of discrimination in the real estate and mortgage lending industries in generating blackwhite residential segregation in metropolitan areas. However, these explanations overlook the role of social institutions in integrating or isolating groups in a community. To address this issue, I develop the closed community thesis and argue that the theological and value orientation of white Conservative Protestant congregations undermines the creation of bridging group ties. Drawing on county-level data from the 2000 Census of Population and Housing and the 2000 Churches and Church Congregations data, I test the hypothesis that the number of Conservative Protestant congregations per 1,000 non-Hispanic whites is directly associated with levels of residential segregation between non-Hispanic blacks and whites. I find that counties with a large number of Conservative Protestant congregations exhibit higher levels of residential segregation along multiple dimensions. Separate models estimated by metropolitan status and region indicate that the effect of Conservative Protestant congregations is consistent across a variety of localities. I conclude by discussing the theoretical implications of the role of religion in explanations of residential segregation and the analysis of segregation outside of metropolitan areas.

Scholars of religion have long known that women are more religious than men, but they disagree about the reasons underlying this difference. Risk preference theory suggests that gender gaps in religiosity are a consequence of men's greater propensity to take risks, and that irreligiosity is analogous to other high-risk behaviors typically associated with young men. Yet, research using risk preference theory has not effectively distinguished those who perceive a risk to irreligiousness from those who do not. In this article, we evaluate risk preference theory. We differentiate those who believe in an afterlife, who perceive a risk to irreligiousness, from nonbelievers who perceive no risk associated with the judgment after death. Using General Social Survey and World Values Survey data, multivariate models test the effects of gender and belief on religiousness. In most religions and nations the gender gap is larger for those who do not believe in an afterlife than for those who do, contradicting the predictions of risk preference theory. The results clearly demonstrate that the risk preference thesis is not a compelling explanation of women's greater average religiosity.