The Audacious Epigone explores the GSS in terms of the relationship between irreligion and politics. He confirms what I’ve noted before, the data shows that the irreligious strongly tend to be liberal, but liberals only weakly tend to be irreligious. This makes sense, the set of liberals in the United States is an order of magnitude more numerous than the set of atheists (2-5% vs. 15-30%, depending on how you measure/quantify). Because most liberals are not atheists, generalizations of atheists do not necessarily apply to liberals. Atheists tend to be more male skewed than theists, while liberals tend to be more female skewed than conservatives.
There are also trends within the GSS in terms of how religion & politics sort out. The variable GOD is rank ordered 1-6 in the GSS, from those who know God does not exist (1) to those who are totally certain (6). POLVIEWS is rank ordered from extreme liberals (1) to extreme conservatives (7). Here are the correlations between these variables taking the numeric equivalents on their face:
GOD-POLVIEWS (blacks) = 0.05
GOD-POLVIEWS = 0.19
GOD-POLVIEWS (whites) = 0.23
GOD-POLVIEWS (whites with college degrees or higher) = 0.33
GOD-POLVIEWS (whites who scored 8 or above on WORDSUM, i.e., higher IQ) = 0.33
As you can see, there’s basically no relationship between religion and politics in the black community in terms of a being able to predict. Most black Americans are Democrats and disproportionately liberal, while at the same time being more religious and theologically conservative than whites. Among whites in the upper socioeconomic strata is where the relationship between religion and politics becomes a trend worth noting.