By Ryan P. Burge, Eastern Illinois University
You never know when a post is going to take off – the internet is funny that way. My last article focused on white evangelicals and their opinion on immigration issues. In it, I describe how white evangelicals are far outside the mainstream on issues about border security, sanctuary cities, and other issues related to immigration. That got picked up by Vox and brought a lot of attention. For most people, I would guess that it confirmed some priors they had about white evangelicals and their view of the world. I try my best to tell data-driven stories. If a group’s public opinions are outside the mainstream, then it’s my job to point that out.
However, the term “extreme” is a pretty loaded one. I will admit to that. The easy retort is “extreme compared to what?” Sometimes we are subject to the anchoring error – that is the first piece of information we hear about something becomes the reference point and everything else is evaluated in light of that initial anchor. So, here’s an exercise in updating our anchor points and putting things in perspective.
Two things I want to accomplish here. First,I wanted to see if two religious groups are as polarized as the parties. Second, I wanted to take it a step further and ask which of those two religious groups is closer to the average American. This exercise was helpful in creating a mental map of where all these groups find themselves in policy space.
I chose two religious groups that are often thought to represent the far ends of the partisan spectrum – white evangelical Protestants (just 16.7% identify as Democrats) and atheists/agnostics (8.2% identify as Republicans). Then I created a sample of all self-identified Republicans, but excluded white evangelicals and another of Democrats that left out atheists/agnostics. That should give us two anchors (where the two parties stand), as well as two groups that can be evaluated against those reference points.
I chose sixteen issues from the 2018 Cooperative Congressional Study that touch on a number of aspects of American life including abortion, trade, taxes, tariffs, Medicare for All, immigration, gun control, and the environment. And because the total sample of the CCES is huge (60,000), the margin of error for each group is very small.
First, note that I coded each question so that the more conservative answer would be further to the right of the graph. As you can see, on almost every issue, white evangelical Protestants are to the left of the average Republican. That means that for all the grief that white evangelicals get for being conservative, they are not as conservative as the Republican sample (not counting white evangelicals). However, in some cases the differences are very small. That will be explored in a minute.
The left side of the graph tells a much different story. Instead of a consistent pattern right down the line, the purple and blue circles alternate back and forth depending on the issue. In effect, atheist/agnostics are more liberal than the average Democrat on seven of the sixteen issues.
If you want to take a look at a zoomed in version of both the right side and left side of the above graph, I visualized that here.
But, let’s get down to brass tacks. If either the nones or white evangelicals could represent the rest of the population in government, which one would be closer? I calculated the mean for the entire sample that wasn’t an atheist or agnostic or white evangelical on the sixteen questions along with the means for both religious groups. In addition, I included the distance between atheists/agnostics and the rest of the sample and put it in the shaded box on the left, and did the same for white evangelicals and the rest of the sample and put it in the right shaded boxed. Finally, I noted which group was closer to the black circle, but for instances when the difference was two percentage points or less, I marked it “no difference.” A hi-res version of this graph is available here.
Here’s the final tally. There were seven issues that span many policy areas where the two religious groups were equidistant from the rest of the sample. These included tariffs on Chinese goods, repealing the ACA, moving the capital of Israel, DACA, and background checks for gun purchases, among others. There were two areas in which the rest of the sample was closer to white evangelicals – both involving taxes. Clearly atheists/agnostics are much more liberal on taxation than the average American. However, on the remaining seven issues, atheists/agnostics were closer to the rest of the sample. The biggest gaps were on withdrawing from the Trans Pacific Partnership, the assault weapon ban, and support for Medicare for All.
It’s fair to conclude that atheists/agnostics do represent the rest of America better than white evangelicals through the lens of these sixteen issues. But, the gaps are so big that neither group represents the public well. On abortion, the gap is 62 points. On moving the capital of Israel, it’s fifty-five points. Even in areas that are not specifically religious, like repealing the ACA, it’s nearly fifty points. The gap on ending the visa lottery is forty-two points. There is no compromise to be found here.
There are a few areas where the distance between left and right is surmountable. On mandatory background checks it’s less than ten points. On DACA, it’s twenty-six points from atheists/agnostics to white evangelicals and it’s just slightly larger than that on an assault weapons ban. Maybe it’s time to seek common ground where it is most likely to be found and then they can get back to yelling about abortion.
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.
If “agree” is supposed to represent the conservative view, then “always allow abortion as a choice” appears to be reversed. Or is the question “Never allow…?” What is the actual wording of the question? If I would allow abortion in all but the most extreme cases, how would I answer?
That’s a good catch. I had to reverse code a few of them so that the liberal responses would end up on the left. So, that one is the % who disagree.