How Has the "God Gap" Changed over Time?

Abstract

For many years, an important feature of the United States’ political landscape has been a strong relationship between religious affiliation and partisan identification. However, recent research has suggested that religiosity is replacing religious affiliation, as a crucial determinant of partisanship. Specifically, religiously observant individuals are more likely to vote Republican than Democratic, a phenomenon referred to as the “God gap.” Furthermore, research suggest that the “God gap” is increasing over time. Using the General Social Survey from 1984 to 2016, I investigate how the relationship between religiosity and partisanship has changed over time. I find that increases in the “God gap” are actually a consequence of an unduly restrictive way of econometrically modeling the passage of time. A specification that allows the data to determine how time affects partisanship suggests that the relationship between partisanship and religiosity has not increased over time. Rather, the positive relationship between religiosity and Republican identification became prevalent in 1998, before which it was insubstantial, and subsequently remains relatively constant.