Abstract: According to the General Social Survey, the combined rate of weekly and monthly attendance at religious services in Canada has declined by about 20 points from 1986 to 2008. About half of this decline stems from the increase in the proportion of people reporting no religion, who, for the most part, do not attend religious services. The other portion of this decline is attributable to eroding attendance rates among Catholics, particularly older Catholics, and Protestants in Québec. Attendance rates for Protestants outside of Québec show signs of increase. The reported increase in weekly attendance in Canada by the Project Canada surveys and cited by Bibby as a possible indicator of a religious renaissance is revealed as an artifact in the data due to an over-sample of Protestants. And a weighting problem with the Canadian Survey of Giving, Volunteering and Participating is identified that leads to under-estimates of religious attendance.

Abstract: In a recent article, Chaves (2010) argues that much of the work in the sociology of religion is susceptible to the religious congruence fallacy-the tendency to assume consistency in an individual’s religious beliefs and behaviors across situations when they are in fact highly variable.  We build and extend this argument by focusing on intersecting group identities as a mechanism for explaining such incongruence, not only within religion but also across other identity categories.  To illustrate this argument, the analysis draws on data from the 2006 Panel Study of American Religion and Ethnicity to assess how race, gender, and religion interact with attitudes and behaviors on key issues of the day, specifically conservative family values and voting behaviors.  The results find marked differences in religion’s association with attitudes and behaviors and inconsistent relationships between attitudes and behaviors across racial-gender groups.  We use the analysis to highlight the conditions that result in incongruence across identity categories and pinpoint where we are most vulnerable to committing the congruence fallacy.  We conclude by offering suggestions for future research.