I've always been a bit fascinated by people who live with inherent tensions in their lives. For example, I reside in a small, rural town that will likely vote overwhelmingly for Trump in 2024—probably by forty points or more. Yet, there are some hardcore liberals here. I know a few of them. They advocate strongly for their beliefs, aware that their efforts might not significantly sway the majority. Many spend considerable time in St. Louis, about an hour's drive away, seeking a cultural environment more aligned with their values.
I’ve written about this before - people who are just in a weird situation. Most notable, I put together a paper about people who both identify as lesbian/gay/bisexual but also say that they are evangelicals on surveys. It was published with the title, “To be of one mind?: integrating an LGB orientation with evangelical beliefs” at Politics, Groups, and Identities. Long story short - these folks are more religious than other LGB people but not quite as active as non-LGB evangelicals.
We have a term for this in political science, it’s called being cross pressured. I consider this piece by Diana Mutz to be one of the foundational works in understanding the concept. The field of religion and politics presents me with a whole bunch of combinations of folks who would clearly fall into this cross pressured category. I wanted to focus on one today that may be the most incongruent - people who identify as atheist or agnostic on the religion question but then say that they are Republicans.
I don’t think I need to provide a huge theoretical justification for why this is a weird combination of factors. The Republican party is basically 85% Christians right now. So to be an atheist who also identifies with the GOP puts you in a really small subset of the population. Let me start by just showing you that - this is the share of Democrats and Republicans who identify as atheist/agnostic over the last several election cycles.
Even back in 2008, atheists and agnostics were much more at home in the Democratic coalition than with the GOP. Twelve percent of all Democrats were atheist/agnostic compared to only 3% of Republicans. Over time, both numbers have increased but the trajectory is completely different. For Democrats, that percentage has risen from 12% to 21% between 2008 and 2022. For Republicans, it’s much more modest - from 3% to 5%. One in twenty Republicans are atheists or agnostics. It’s one in five Democrats. There are four Democrat atheist/agnostics for every Republican.
I think part of the reason for this finding is not the fact that Republicans are just a whole lot more religious than Democrats, it’s that Republicans just don’t like that atheist/agnostic label that much. So, they may be functionally non-religious but they would never want to call themselves a term that they believe to be repugnant like atheist or agnostic. The empirical evidence for that is pretty clear when you limit the sample to people who report their religious attendance as seldom or never and then calculate the religious composition of those low attenders.
For Democrats who attend less than once a year, a majority are nones - 35% are atheists/agnostics and nearly the same share are nothing in particular. That percentage has grown by about ten percentage points since 2008. Only 27% of never/seldom attending Democrats say that they are Protestant or Catholic today.
However, those same figures for Republicans who are low attenders are much different. Among those who are seldom/never attenders, 56% of them say that they are Protestant or Catholic - that’s twice the share as the Democrats. Meanwhile, only 11% of low attending Republicans are atheist/agnostic - compared to 35% of Democrats. Pretty strong evidence here that when Democrats are far from religion, they have little hesitancy in embracing the atheist/agnostic label. Not so for Republicans.
But let me take that a step further and add another question to the mix - religious importance. I calculated the share of Democrats and Republicans who said that they were atheist/agnostic broken down by level of religious attendance and religious importance. This is where my thinking crystallizes quite a bit.
Obviously there’s not much happening in the top right of these heatmaps - those are folks who score both high on religious importance and religious attendance. Almost none of them say that they are atheist/agnostic. The real action is happening on the left side of the graph and the color coding tells a lot of the story. Look at the bottom left square, specifically. Those are people who never attend religious services and say that religion is not at all important. Among Republicans, 33% of those in this box are atheist/agnostic, it’s 58% of Democrats.
Those huge gaps are evident across a number of combinations of importance and attendance. For instance, among those Democrats who say that they seldom attend religious services and religion is not at all important - 40% are atheists/agnostics. It’s only 24% of Republicans in that same square. In fact, there’s not a single combination of these two factors in which a Republican is more likely than a Democrat to identify as atheist/agnostic.
Okay - I think it’s pretty well established that even if you control for other questions related to religion, a Republican is just less likely to identify as an atheist or agnostic compared to a Democrat. There’s some clear hesitancy there. But I was curious about something else - are Republican atheists/agnostics that different politically than the Republican party as a whole? And, just for the fun of it, I did the same thing for the Democrats. The area of inquiry was questions about abortion. That seems like a topic that could be deeply impacted by religious convictions, so I wanted to see if an atheist Republican was less conservative than the average member of the GOP.
The data on this is pretty convincing, atheist/agnostic Republicans are more moderate on abortion across a variety of scenarios. For instance, while only 30% of all Republicans would favor a woman having access to an abortion for any reason, that share rises to 51% of atheist/agnostic Republicans. Only ten percent of the latter group favors a complete ban on abortion, compared to 25% of Republicans in general. On the bottom two abortion questions, there’s a thirteen point gap. Atheist/agnostic Republicans are less willing to support a ban on late term abortions and less willing to favor a prohibition on federal funds for abortion.
There are gaps among the Democrats, too, by the way. Atheist/agnostics here are consistently more pro-choice than the party as a whole. You can see that especially on the question about late term abortions. About a third of all Democrats would support a ban after 20 weeks of gestation, it was only 13% of atheist/agnostic Democrats. My guess is that a lot of Black Protestants and Hispanic Catholics in the larger Democratic sample are skewing the numbers a bit there.
But do atheist/agnostic Republicans actually vote any differently than Republicans as a whole? Yeah, they do. That’s what is coming through when I calculated the vote choice of those two groups over the last four presidential elections. The gaps aren’t huge, but they are certainly noticeable.
For instance, McCain only got 82% of the atheist/agnostic Republican vote in 2008. But, Romney did a whole lot better in 2012. His percentage jumped up to 89%. However, Trump did really poorly with this group in his first campaign in 2016. He only got 80% of Republicans who identified as atheist/agnostic. It’s interesting that the remaining 20% was fairly evenly split between Clinton (12%) and third party candidates (8%). However, in 2020, Trump did much better with this group - getting back to near Romney levels. This may have something to do with the fact that third party candidates were not as viable in that election cycle.
Here’s what I know - there are going to be more atheist/agnostics who identify as Republicans in the future. It’s almost inevitable at this point. A group like the nones can’t get to 30% of the population by just drawing from the same segments of society over and over again. It will have to get more politically diverse in order for it to grow. How can the GOP be hospitable to this group, while still remaining the party of a whole of evangelicals? Time will tell.
I do think that Trump’s posturing on abortion is probably a good strategy in this regard, for what it’s worth. Making it a state’s rights issue is a way to sidestep the larger moral questions at the federal level and letting voters decide in those states is probably a pathway forward that doesn’t turn off many Christian conservatives and likely doesn’t repel the growing number of Republican nones.
Code for this post can be found here.
Point of personal privilege here.
Our family got a dog in October of 2009. She was a little Yorkshire Terrier named Lucy. She never weighed more than seven pounds. She also never left my side. I say this without any exaggeration - that dog was laying right next to me in my office chair for every word I’ve written during the the last dozen or so years. She was there when I wrote my dissertation, when I signed my first book contract, and when I started this Substack.
She moved in with my wife and I about ten weeks after we bought our house. She was there when we renovated nearly every part of it. She was waiting when we brought home our two boys from the hospital. It’s amazing how much one’s life changes between 27 and 42 years old.
Yorkies are attachment dogs and I was her person. She demanded that I take her on two or three walks every day. As she moved into senior status, those walks got shorter, her pace get slower, her world got smaller. I acted like I hated taking her around the yard. I really did love it. It was a perfect opportunity to stand up, walk around, and think about what I was writing before I started tapping the keys again. She was the physical manifestation of the idea that we are not built to just work. We need breaks, too.
Lucy’s time on this Earth ended on Saturday, May 4. She would have been 15 years old in August. Every day of her life I told her two simple things:
You are the best dog they ever made. You are going to live forever.
I knew both of those statements were false, but I very desperately wanted them to be true.
I’m going to miss that dog.
I am so sorry for the loss of your dog Lucy. I'm still grieving my beagle, to whom we said goodbye last year. It hurts.
So sorry for your loss. Glad you got to be her human!