Provo, Utah, and Burlington, Vermont, represent opposite ends of the U.S. religiosity spectrum.
Provo, Utah, tops the list of America's most religious metros, according to survey results recently released by the Gallup Organization. More than three-quarters of residents in this metro reported that they are "very religious." Three of the top five most religious metros are in Alabama — Montgomery, Birmingham, and Huntsville. Jackson, Mississippi, also ranks among the top five.
Burlington, Vermont, and Boulder, Colorado, tie for the title of the nation's least religious metro area. In these two metros, fewer than one in five residents claim to be "very religious." San Francisco and Boston also rank among the least religious metros in America, as only about one in four of their residents say they are "very religious."
The data are based on the Gallup-Healthways Well-Being Index survey of nearly 250,000 Americans across 189 metros conducted over the course of 2012.
Religion in America has an unmistakable geographic dimension, as the map below (based on the Gallup-Healthways data) by Zara Matheson of the Martin Prosperity Institute shows. The most religious metros as well as states are in the South, while the least religious are in in the Northeast and West Coast. As Gallup points out: "The most and least religious cities generally reflect the religiousness of the state in which they are located, although there are some interesting exceptions ... America is a remarkably religiously diverse nation, and much of this diversity is geographically based. Residents in some areas and cities — namely, those in the South and in Utah — are two or three times as likely to be very religious as those living in cities in the Northeast, the Northwest, and other Western locations."
With the help of MPI's Charlotta Mellander, I took a look at some of the key factors that might be correlated with religiosity, measured as the share of respondents who identified as very religious. As usual, I note that correlation only points to statistical associations between variables; it does not imply what causes what.
For one, there is a close connection between religion and political views. Religiosity is positively associated with the percent of metro residents that voted for Romney in 2012 (.60, the strongest correlation of any in our analysis), and negatively associated with the percent of residents who voted for Obama (-.59).
Religion is closely associated with socio-economic class. Metros with lower incomes, lower levels of college grads, and working class economies are more religious. Religiosity is positively associated with the share of working class jobs (.49), and negatively associated metro income levels (-.55), the share of adults that are college graduates (-.42), the level of innovation (-.45), and the share of knowledge, professional and creative jobs (-.35).
Religion is also associated with two markers of demographic diversity, being negatively associated with the share of metro population that are immigrants (-.31) and gay (-.44).
This is in line with my previous analysis here on Cities which found religiosity at the state level to be positively associated with political conservatism, poverty and working class economic structures, and negatively associated with political liberalism, knowledge workers, and college grads.
As I pointed out then, these patterns conform to the broader findings "of political scientist Ronald Inglehart, whose detailed World Values Surveys identify the shift from religious to secular values as one part of the transition to more economically advanced societies. Politicos on the left and right like to explain religious voters' proclivity purely in terms of values. But this misses a central point - that religion is inextricably bound up with the nation's underlying economic and geographic class divide."