(ANALYSIS) I can’t point to one specific instance of this, but it’s something I see in the online discourse:

“It’s the folks who don’t go to church who put Trump in the White House.”

“It’s the most religiously devout Republicans who are the ones driving the MAGA train”

Which one is right? Both are. Which one is wrong? Both are.

It’s actually a really complicated question that interacts three major forces in American life: religious attendance, race and political partisanship.

This is going to get a bit wonky, but hopefully you can follow along here. There’s also going to be a lot of graphs along the way. So, buckle up.

Let’s start at a pretty high level. Were folks who attend church more frequently the ones who voted for Donald Trump at the highest rate? Without a doubt, the answer to that question is yes! But there are lots and lots of caveats to this discussion.

The first caveat is that the relationship between attendance and a Republican vote has basically looked the same in the previous four presidential elections. When Donald Trump won the Republican nomination for president in 2016, one of the biggest questions was if religiously active Republicans would support him.

A better question would have been whether we should assume that the candidate actually has any impact on the voting patterns of Americans. In “20 Myths,” I wrote an entire chapter about this. Joe Biden did a little *worse* among White Catholics than Hillary Clinton four years earlier. And the guy goes to Mass every weekend!

Thus, it’s helpful to go into any election with the assumption that not much is going to change in how folks vote every four years. Very religiously active folks are more likely to vote for Republicans. That was true with McCain in 2008. It was true with Trump in 2020.

Just to put a fine point on this, here’s the Republican share among White voters from never attenders to weekly+ attenders.

2008: 29% → 82% (+53)

2012: 30% → 82% (+52)

2016: 38% → 82% (+44)

2020: 34% → 85% (+51)

So, 2016 was a bit of an outlier but not by a ton. And really that was more about never attenders being slightly warmer toward Trump than anything else.

But notice something subtle I did there? I was only discussing White voters, not people of color. And when you bring them into the conversation, the narrative changes a whole bunch.

While more religiously active people of color were more likely to vote for the Republican, the pew gap is just not that big.

Here’s the pew gap (percent vote for the GOP among weekly+ attenders minus the percent vote for the GOP among never attenders) for White voters versus people of color:

2008: 53 points, 20 points

2012: 52 points, 15 points

2016: 44 points, 11 points

2020: 51 points, 13 points

When we talk about religious attendance and voting patterns, we really need to put a racial qualifier on there. There’s just a huge difference between White and non-White Americans on religious attendance and the ballot and it can’t be ignored.

Let’s shift the focus a bit now and look just at Donald Trump — specifically, his approval rating in 2020. I like approval rating because it doesn’t map exactly onto vote choice. You can dislike someone but still vote for them. So, this is Trump’s approval in late 2020 broken down by religious attendance and race.

That same big gap shows up, without a doubt. For White voters, approval runs from 34% among never attenders to 80% among White people who attend church multiple times per week. Among voters of color, it’s a different story, with 19% of never attenders approving of Trump; it was 32% of those who attended more than once a week.

Here’s a staggering stat for you: A never-attending White person was more likely to approve of Trump in 2020 than a person of color who attended religious services multiple times a week.

Let’s add one more layer to this. We have already considered race and religious attendance, but now I am going to add political partisanship. Again, we are considering Trump’s approval rating in November of 2020.

Guess who loves Donald Trump? Republicans. Especially White Republicans. His approval rating was 90% of White Republicans who never attend. It was 95% among White Republicans who attend more than once a week. But that’s basically the case among non-White Republicans, too. Approval bounces around a bit more, but it averages out to basically the same number: ~80%.

Among Democrats, approval is predictably low. Among never attending White Democrats, Trump’s approval rating was 3%. It does go up among weekly attenders to 12%. (Let’s ignore those weekly+ White Democrats — they literally make up .5% of the sample.) Overall, approval is very low and stays low, regardless of attendance.

The same is true for Democrats who are people of color. No matter how much (or little) they attend services, very few approved of Donald Trump. The share bounces around from 4% to 10%. But, again, there’s no magical effect that church attendance is having here.

Do you see the point I am trying to make? It’s not church attendance that’s doing the work here. It’s partisanship. Church attendance is really a proxy for partisanship.

This is what we call an endogeneity problem. That’s a fancy word, I know. We also call it circular causation, too. But here’s a description that probably makes more sense to the average readers: It’s a chicken or egg problem. Which came first?

In this case:

Does being a Republican make someone more likely to go to church?

Or does going to church more often make someone more likely to be a Republican?

There’s an assumption out there that evangelical churches are indoctrination factories just pumping up the Republican rhetoric. I just don’t think that’s true. And I think that the data points to the same conclusion.

Here’s what I mean by that. I took the sample and broke it down into Democrats and Republicans. Then, I subdivided it by religious attendance.

I wanted to look at issue positions here, not just vote choice on election day. Voting doesn’t perfectly predict issues positions; there may be some “play in the joints” here. For instance, Republicans who don’t love guns or Democrats who aren’t big fans of abortion.

Guess what? Religious attendance just doesn’t matter a whole heck of a lot once you control for partisanship.

I do think there’s an argument to be made that religious attendance actually matters for Democrats. It pushes them in a more conservative direction, really. For instance, when a Democrat attends more often, they are more likely to want to ban abortion completely, cut legal immigration by 50% and be less inclined to ban assault rifles. So, churchgoing Democrats are more conservative than those who don’t go to church.

What about Republicans? Not much to report there. A Republican who attends weekly+ is just as likely to support cutting legal immigration as one who never attends. They have the same view of racial problems and banning assault rifles. The only real difference is on making abortion completely illegal. Among never-attending Republicans, it’s 15% in favor. Among those who attend multiple times a week, it’s 60%.

But, for most issues, partisanship is way more important than religious attendance. That’s particularly true for Republicans. No matter what’s happening in those Sunday sermons, it’s not changing minds on immigration, race or guns (in either direction).

Now, there are two reasons why that could be the case. One is that pastors just don’t talk about those types of issues from the pulpit. I made that argument in the Wall Street Journal last year. The other explanation is that pastors are trying to change people’s minds, but they are just failing. And if that’s the case, then there should be a lot of religious leaders who need to do a lot of soul searching about the influence they have on their congregations. From this vantage point, it doesn’t look like they have much.

Before I finish, I wanted to make this super complicated: race + attendance + partisanship. The same four issue areas that we previously discussed: abortion, immigration, guns, and race. Does race play a big role here?

For White Democrats, religious attendance drives up conservative policy positions in all four areas. White religion pushes Democrats to become Republicans. That’s a spicy take.

For non-White Democrats, it’s a bit more of a nuanced story. Attendance drives up support for an abortion ban and cutting legal immigration. But it drives down the belief that racial problems are rare in the United States. Definitely a mixed bag there.

For White Republicans, attendance just doesn’t matter — except when it comes to abortion, then it matters a whole lot. For non-White Republicans, it’s basically the same pattern. On issue areas, there’s just not a whole of movement outside of abortion.

But the upshot is this: I just don’t think religion, as it’s currently being measured, is that impactful when it comes to politics. Religion now lives downstream of politics. What I mean by that is people don’t go to church to have their political views challenged. They just don’t. They look for a church that either (a) doesn’t talk about politics at all or (b) only talks about it a way that affirms their beliefs.

This really should lead to some real introspection among pastors, priests, imams and rabbis. What impact are we actually having on the worldviews of the folks who come to worship each weekend? To put on my pastoral voice for a second, “Who is discipling these people when it comes to politics?”

It doesn’t look like it’s religious leaders.

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. Subscribe to his “Graphs about Religion” column on Substack, where this post first appeared.