By Ryan P. Burge, Eastern Illinois University

I’ve written about it before, but it’s a data point worth returning to – not only has American politics become polarized, but that polarization has seeped into American religion as well. Christians traditions, by and large, have drifted toward the Republicans. The reasons for this would fill hundreds of pages. Some of them were deliberate – denominational leadership intentionally moved the theology of their churches in a conservative direction. But regardless of the reasons, American Christianity has become increasingly tied to the GOP. The religious “nones”, on the other hand, seem to be a group that is more liberal politically. This would bring some type of balance to American politics — ~23% of people are evangelicals (who overwhelmingly vote for Republicans) and another ~23% are nones (who strongly support Democrats). But that leaves out over half the population. What are all the rest doing?

To simplify my argument I wanted to just look at three groups of people – white evangelicals, white people who have no religious affiliation, and the rest of white people who are neither evangelical nor none. I turned to the General Social Survey from 1972-2018 and I am using political partisanship as a simple proxy for political polarization.

The trends are indicated by LOESS lines for each group, with standard errors visualized by the shaded ribbons. As can be seen on the left side of the graph, the nones and evangelicals were politically indistinguishable from 1972 to 1980. That’s worth noting. However, from that point forward we can see that the lines diverge. Evangelicals begin moving fairly rapidly to the Republican side of the spectrum, while the nones’ line stays fairly flat. This rapid move for evangelicals lasted until the mid-1990’s. At that point, evangelicals still drifted toward the Republican side, just at a slower pace. The nones also saw some changes after the mid-1990s. They slowly began drifting toward the Democratic side of the continuum. However, this move was slower and more subtle than it was for evangelicals.

So, where does the average white respondent fit between these religious groups? I plotted that below with a black LOESS line. Again, until 1980 there are very small differences between the three groups. However, between 1980 and the mid-1990’s the white sample begins to drift somewhat toward the Republican side. Yet, this is where the movement stops for white people who don’t identify as nones or evangelicals. The data clearly indicates that this group of white people has the same political partisanship in 2018? that they had in 1990. That’s a stunning amount of consistency and a clear repudiation that we are a “center right country” (instead, it appears to be center-left and this is just among whites).

So, have evangelicals moved further away from the center or was it the nones? While it’s impossible to know what the “middle” of American politics looks like, I think it’s not much of a leap to say that the black line described above can serve as a benchmark for political moderates. So, using that line as my reference I calculated the distance from evangelicals to the benchmark, and the nones to the same reference point. I indicate on the graph three points along the way – 1990, 2006, and 2018. I picked these arbitrarily but I don’t believe them to be outliers. The percentages represented below are differences between the two lines. The scale runs from 0 (strong Democrat) to 6 (strong Republican), so a difference of .6 is 10%, for instance.

In 1990, the average (non-evangelical/non-none) white person was 5.3% away from white evangelicals and 5.4% away from religious nones on the political spectrum. The differences are a statistical wash, and it’s also symmetrical polarization. However, once we move to 2006 things have changed. There the average evangelical is now 14.1% away from the average white person, while the distance to the nones is 4.7%. In effect, the nones’ shift looks like the shift in the rest of the white sample (a little to the left), while evangelicals had nearly tripled their (1990) distance away from the typical white respondent. The results from 2018 bring some nuance to the conversation. Now, the average evangelical is 15.2% way from the average white person and still drifting away from the center. While on the other hand, the average none has now shifted left somewhat – with the gap widening to 8.3%. In effect, evangelicals are now nearly twice as far away from the midpoint as the nones.

It’s a clear case of asymmetrical polarization. Yes, nones have moved to the left. But that move was from 5.4% to 8.3%. The shift for white evangelicals was from 5.3% to 15.2%. In essence, evangelicals have increased the distance between themselves and the benchmark by 300%, while nones have increased that distance by about 50%.

There are lots of caveats to all this:

1. I am fully aware that there are mainline Protestant churches in the United States that are more moderate (I happen to be one of them). However, they are disappearing at a rapid rate. The average mainline Protestant is nearly 60 years old, and many of them aren’t even that moderate.

2. I don’t know how well that the average white non-evangelical/non-none can perceive this shift. In a toxic and chaotic political landscape where parties hurl inappropriate and inaccurate labels at the other, and the media relies on extreme views to drive clicks and viewership – it’s difficult for the average person to know what has actually occurred in American politics and religion over the last forty years. It may be impossible for the average person to know what the religious composition of the parties looks like.

3. We just don’t have enough data to calculate just how many Americans may have left evangelical congregations over political grievances. Although Djupe, Neiheisel and Sokhey found some evidence of it in during the 2016 presidential election, there’s just not survey data for other time periods.

I think that the headline for an article that I wrote for Religion News Service sums it up well: “Why politics may kill white churches.” While becoming more politically extreme may build a moat around evangelical Christianity, its adherents need to remember that moats are good at keeping people from getting out, but it makes it much more difficult for people to get in. The politics of white evangelicalism may provide a floatation device in 2018, but it could become an anchor in the coming decade.

Ryan P. Burge is an assistant professor of political science 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.