(ANALYSIS) I was thinking a bit today about the idea of subgroup composition. For instance, evangelicals could be the same share of the population today as they were in the early 1980s, but that doesn’t mean that the composition of the group hasn’t changed significantly during the prior four decades. In fact, it would be pretty shocking if the racial composition of evangelicals hadn’t shifted and the average educational attainment hadn’t climbed, given the overall macro-level movement in American society.

That got me thinking quite a bit about a specific group — those who never attend religious services. In 2008, according to the Cooperative Election Study, about 20% of all respondents reported that they never attended religious services. By 2022, that share had risen to 34%. A 14 point jump is a whole lot of folks, by the way. In fact, in real numbers that’s over 45 million Americans.

But the composition of never attenders has also changed as that group has grown so much larger. What I really wanted to do is help readers better conceptualize this group — especially when it comes to politics. One of my hobby horses recently has been trying to convince people that they need to stop thinking about Republicans as incredibly religiously active and Democrats as the ones who have nothing to do with religion. The Republican coalition is looking less and less religious every year, and this is going to have big impacts in the elections to come.

Let’s start broad — the share of each party identification that never attended religious services between 2008 and 2022.

In 2008, independents were the most likely to be never attenders — bet you wouldn’t have guessed that. Twenty-eight percent of them checked the “never” box, which was four points higher than Democrats. There were very few Republicans who were never attenders back when Obama faced off against McCain for the White House — just 10 percent.

But all those figures have risen significantly over the last 14 years.

In 2022, the share of independents and Democrats who were never attenders rose to 42% each. That was a jump of 18 percentage points for Democrats and 14 points for independents. Among Republicans, the share had also risen — 22%. More than double the rate of 2008.

But here’s where things get really interesting — I wanted to just look at the partisan composition of never attenders.

In 2008, 62% of never attenders were Democrats and 20% were Republicans. That’s a total gap of 42 points.

In 2012, that gap was essentially the same. But in 2016, it dropped to only 31 points (56% vs 25%). It stuck there for the 2020 election as well. Those percentages didn’t change at all.

Then, in 2022, the gap was just 29 points (53% vs 24%) — 42 points in 2008, 29 points in 2022. The Democrats are losing their advantage with never attenders.

But, where are those Republican gains coming from? Older folks? Younger ones? Let’s find out. This is just the sample who never attended religious services in 2008-2010 compared to 2020-2022 (I grouped these years together to give me a bigger sample size). I did this by birth year so we can more easily compare when the gap opens up.

The answer is clearly older folks. Among people born around 1950 who never attended religious services, about 27% were Republicans in 2008-2010. A decade later that share had jumped to 34%.

That gap is fairly large for people born between 1940 and 1955 — then it begins to narrow significantly.

Among people born in the 1970s and early 1980s, there’s really been no major partisan shift among never attenders. However, among those born in 1990 (they were just coming into adulthood in those early surveys), the gap opens back up.

Just 12% were Republicans in 2008-2010, compared to 17% in the last few years.

But here’s where things get really odd. Partisanship is one thing, but vote choice is another measurement entirely. This is the vote pattern of never attenders over the last four presidential elections.

Obama had a huge advantage with this group when he won the White House in 2008. He earned 72% of never attenders, compared to only 26% for McCain.

He managed to lose some ground in 2012 — losing four percentage points.

But Hillary Clinton did a lot of worse.

Her advantage among never attenders was just 27 points (60% vs 33%). In 2008, Obama’s margin was 46 points. This is one of the reasons she lost, I suspect.

By 2020, Biden seemed to reverse the trend. He basically got never attenders back to the Obama reelection level. He took 69% of the vote, and Trump only managed 28%.

Part of this was the lack of a viable third party candidate, for certain. But that doesn’t tell all of the story because that was only 7% in 2016.

Let’s go even further now. This is the ideology (liberal, moderate, conservative) and partisanship (Democrat, independent, Republican) of never attenders in the last four presidential election cycles.

Forty-three percent of never attenders were liberal Democrats in 2008, and another 20% were moderate Democrats. That’s basically two-thirds of never attenders in those two squares. But 14% were conservative Republicans, as well.

In 2012, the share who were liberal Democrats dropped to 37%, and the conservative Republican bucket went up by two points.

In 2016, only 35% of never attenders were liberal Democrats — down eight points from 2008. In total, 54% of the sample was in the bottom left two squares, compared to 63% eight years earlier.

Definitely some evidence that the never attenders category was more moderate by 2016. But in 2020, the entire trend reverses itself. Now, 43% of never attenders are liberal Democrats — that’s exactly the same share as it was in 2008.

However, the share of moderate Democrats was just 15% in 2020 compared to 20% in 2008. There were more moderate independents in 2020 compared to 2008, and the conservative Republican never attenders also grew during this time period.

I wish there was a clear direction in these results, but there just isn’t. Never attenders started out as pretty liberal then began to drift back to the middle of political and ideological spectrum. But then, in 2020, they turned back harder to the left, and never attenders were just as likely to be liberal as they were in 2008.

So, that group grew by 14 percentage points, but the overall ideology/partisanship didn’t really change.

Here’s where the clouds part for me a bit, though. I put together a regression model. The dependent variable (the thing I am trying to predict) is whether someone is a never attender. I wanted to see if the same factors worked the same way in 2008 compared to 2020 to predict someone never attending religious services.

Here’s how to interpret this — anything to the left of zero means someone is LESS likely to be a never attender. Anything to the right predicts a GREATER likelihood of never attending. If it overlaps with zero, there’s no relationship between that variable and never attending. (By the way, these are robust standard errors, and the coefficients are scaled, which means they can be directly compared).

What predicts a greater likelihood of never attending? Being a man and being White. That’s it. What predicts a lower likelihood of never attending? Age, income, education, being a Republican and being an evangelical.

That group of variables should come as no surprise to anyone who has been reading my work, by the way. Easily the most predictive is identifying as an evangelical, but second in line is being a Republican.

But here’s what I am going to take from this — notice how in many cases the estimate from the 2020 model is closer to zero than the estimate from 2008? That’s the case for evangelical ID, Republican and male. You know what that means? Those variables have less explanatory power now than they used to have. In layman’s terms, that indicates that a never attending response is less unique than it used to be.

It used to be easier to predict whether someone was a never attender by whether they identified as an evangelical or a Republican or a man. Now, those variables still have explanatory power, it’s just less. Never attenders are starting to look more and more like the general population — not some unique group like they were in 2008.

This is what’s going to happen as a group gets to a third of the population, by the way. This is what I always say about the nones — you can’t get that large without drawing in people from basically every demographic group. Young and old. White people and people of color. Republicans and Democrats. That’s what is happening among never attenders. It used to be more Democrats and more young folks. Now it’s just a whole lot of everyone.

This piece is republished from Graphs About Religion on Substack.

Ryan Burge is an assistant professor of political science at Eastern Illinois University, a pastor in the American Baptist Church and the co-founder and frequent contributor to Religion in Public, a forum for scholars of religion and politics to make their work accessible to a more general audience. His research focuses on the intersection of religiosity and political behavior, especially in the U.S. Follow him on Twitter at @ryanburge.