by Ryan P. Burge, Eastern Illinois University 

Every once in awhile I will post a graph that has a category for evangelicals who never attend church. And it’s inevitable that I get the following response:

I get the impulse here. Most people think of religion as both something you do (like attend church) and something you believe (heaven and hell, literal Bible, etc.) However, that’s not the primary way that I think of American religion. I believe that choosing a religious affiliation on a survey is an exercise in placing oneself in social space. If I never attend church but was raised evangelical, a survey question about religious affiliation puts me in an interesting spot. I could just as easily say that I’m “nothing in particular” as I am to say Southern Baptist.

But when someone chooses Southern Baptist when given all those other options, they are making a conscious choice to say that Southern Baptists are “people like me” (to use a phrase from Achen and Bartels). Taking on that identity means something. I think it means I still hitch my wagon to the Southern Baptists in some meaningful way – that we are on the same team. And the data backs me up on that.

Let’s start by just getting a lay of the land. What percentage of each religious tradition says that they never attend church services? It varies widely, as you can probably guess. For instance, about one in twenty Mormons say that they never attend services. That’s about the same as non-white evangelicals. In comparison 9 percent of white evangelicals never attend church. It’s a bit higher for non-white Catholics (12.4%) and white Catholics (14.7%). And a quarter of Jews never attend church.

The top of the graph is no surprise – these three groups (atheists, agnostics, and nothing in particulars) are considered to be the “nones”. But, note that there are pretty significant disparities in how often each doesn’t attend church. Only half of nothing in particulars never go. It’s seven in ten agnostics, and nearly 90% of atheists.

However, do lower levels of religious behavior correlate with lower levels of religious belief? The CCES asked respondents: How important is religion in your life? This is helpful because it is not specific to any religious tradition. One would assume that people who never go to church also think that religion is not that important in their life. But, that’s not what the data says.

Note that the groups comprising the nones are the most likely to say that religion is not at all important. That’s true for 97.2% of atheists, 84.7% agnostics, and 61.3% of nothing in particulars. That makes some sense. However, once we move into a number of theistic groups some significant differences emerge.

For instance, half of all Jews who never go to services say that the religion is not at all important. That’s true for just a third of all Muslims who never attend. But, all the Christian traditions in the sample are much less likely to say that religion is not at all important.

About 3 in ten white Catholics say that it’s not at all important, it’s 26.5% of non-white Catholics. Mainline Protestants are just slightly higher at 28.9%. But there’s a big drop when moving to evangelicals. One in ten non-white evangelicals who never attend say that religion is not at all important, it’s slightly higher for white evangelicals (14.6%). It’s clear from this angle that never attenders are not created equally. Many of them still put a great deal of emphasis on their religion, even if they never go to church.

But, does this make any difference in modeling? To test that I specified a simple logit regression, predicting the likelihood of someone identifying as a Republican. I limited the sample to just those who never attended church. I added a number of controls for basic demographics that change the likelihood of association with the GOP. These included variables for age, education, income, gender, and those living in the South. The interpretation is straightforward – point estimates to the right of the dashed line equate to a greater likelihood of identifying as a Republican, those to the left are a lower likelihood. If either the point estimate or the horizontal blue line intersect with zero, then that variable does not have a statistically significant effect on identifying as a Republican.

The demographics work as expected: older folks, men, and those living in the South are more likely to identify with the GOP. Higher levels of education were negatively related to Republican affiliation. But speaking in terms of religious tradition, a never attending evangelical is more likely to identify as a Republican than mainline Protestants and Catholics (nones were the reference group here). In short, people who identify as evangelicals despite never attending church are more closely linked to the GOP than other nominal Protestant identifiers even after controlling for key demographics.

The model bolsters my central assertion – choosing to identify as an evangelical, even among those who never attend church, sends an important signal about how you see yourself and who you see as “your people.” As such, religious affiliation moves beyond the realm of theology and becomes both a social and political construct. So, once and for all, you can be a never attending evangelical and it matters.

Featured image from here.

Ryan P. Burge teaches at Eastern Illinois University in Charleston, Illinois. He can be contacted via Twitter or his personal website. The syntax for this post can be found here.