Highly Educated Men Are More Likely to be in Church Than Highly Educated Women

Is it partisanship? Does this filter into religious composition, too?

Let me get this out of the way right up front - this post probably hangs together a bit less than other things I have written because it was more of a “stream of consciousness” in terms of graphs than anything else. But I saw something in a piece of analysis that I couldn’t stop thinking about and wanted to make more sense of it by chasing some loose threads.

Here’s the setup - educated people are more likely to attend religious services weekly than those with a lower level of education. I wrote a long piece about that a while ago:

But someone (I can’t recall now, it was maybe in a Twitter reply) asked if that same relationship held for both men and women. And so I put together this piece of analysis. It’s just a simple graph of weekly attendance, broken down by level of education and gender. I tested it across multiple waves of the Cooperative Election Study. Here’s what I get.

In 2008, it’s a positive relationship for both genders - higher education leads to higher attendance. But in 2012, the shape of the trend lines begins to look different. For women, there’s really no appreciable benefit of education on church attendance. About 30% of women with a high school diploma or less attended weekly, it was 32% of those with a college degree. For men, that gain is pretty large at seven percentage points. Education gets men to church. It doesn’t for women.

In 2016, that same finding reappears. For men, the weekly attendance rate jumps eight points. It only rises two points for women. It’s there in 2020 (22% to 30% for men. 24% to 27% for women). It’s an eleven point gain for men in 2022 and 2023. The increase for women is just three percentage points.

Education still drives weekly attendance up, but the effect for men is three times larger than it is for women. That’s…something.

Let’s go on a bit of a journey here where I try to pull apart several variables of interest - gender, education, partisanship, and religious attendance to see if I can make any sense of this disparity between men and women.

I broke the sample down into six subgroups across three levels of education and both genders and then tracked the partisan composition of each group from 2008 through 2023.

You can pretty much see the growth of education polarization in this graph. Compare men with a high school diploma or less versus those with a college degree. In 2008, about the same share were Democrats (~42%). Today about 45% of college educated men are Democrats compared to only 35% of those with a high school diploma. But notice among the highly educated, the Democrats clearly began to outnumber the Republicans around 2014 or so. Now that group is D+6 when it used to be R+2.

The trend lines for women just look so completely different than men. Look at the bottom left graph (women with a high school diploma or less). In 2008, 50% of this group were Democrats and about 33% were Republicans. Today, 40% are Republicans and 37% are Democrats. There has also been a big drop among women with some college education. About 52% were Democrats in 2008, now it’s closer to 45%. Meanwhile the Republican line hasn’t changed much.

But women with a college degree look like no other group in this graph. About 60% were Democrats in 2008. It’s dropped just a bit to 57% in the most recent data. The Republican share has dropped by about the same amount.

A college educated women is 10-12 points more likely to be a Democrat than a man at the same level of education. It’s fair to say that the Democrats strongest base are women with a college degree. They do have an advantage with highly educated men, but it’s much smaller.

Let’s inject religion into the equation now. This is four categories - women/men, Republican/Democrat. I calculated the share who were Christians and nones since 2010.

Among Democrat males, the nones were in equal number to the Christians somewhere around 2018 - about 45% each. The lines began to separate a bit more by 2020 but then they were converging again. In 2023, the share of male Democrats who were Christians was the same as the share who nones.

For female Democrats, the lines still haven’t crossed. The gap in these two numbers was huge in 2008 - 60% were Christians and 25% were nones. Today, it’s less than five percentage points. But the trend lines have flattened out, too. So this convergence is not a foregone conclusion.

For Republicans, it’s just a whole different ball game. Over 75% of men were Christians in 2010 and 15% were nones. Those figures for women were close to the same - 80% vs 12%. Now, it’s important to note that the Christian share is dropping for both Republican men and women. Today about 68% of Republican men are Christians and about 70% of Republican women. But the gap here is still fifty points for Republican women, it’s five points for Democrat women.

I am still trying to square this finding that education works different for women than it does for men when it comes to religious attendance. Maybe this is a generation thing. Young educated women are just way different than young educated men when it comes to their religious affiliation. So here’s the religious breakdown of younger educated men vs younger educated women over the last few years.

Now, I’m just baffled a bit. Among men in 2008, 56% were Christians and 32% were nones. Among women, 65% were Christians and 27% were nones. Okay, pretty good evidence that young educated women were more religious than men in the same category. But look at 2020. Among men, 49% were Christians and 40% were nones. For women, those figures are 49% and 41%, respectively. That’s absolutely nothing to write home about.

Now, I will say this: I think it’s fair to conclude that the composition of the nones among college educated younger people is different based on gender. When men become nones, they are much more likely to pick the atheist or agnostic category. For instance, about a quarter of all men with a college degree between the ages of 18 and 40 were atheists or agnostics in 2020 and 2022. For women it was about five points lower. Meanwhile, only 15% of men were nothing in particular. It was about six points higher among women. Atheism is just not that attractive to younger women.

But let’s bring this full circle and get back to the question of education, gender, and religious attendance now. I put together a regression model, predicting the share who were weekly attenders by gender and education. I controlled for a bunch of stuff: age, income, race, marital status, parental status, and political ideology.

Among those with a high school diploma, all things being held constant, a woman is slightly more likely to be a weekly attender compared to a man. But the increase is small (~2%). For folks with a two year degree, it’s about the same story - 22% for women and 20% for men.

But here’s what is fascinating - the last two sets of bar graphs. Among those with a four year college degree, there is absolutely no difference in religious attendance based on gender. About 24% of women and men attend weekly once you control for all kinds of demographic factors. The last set of bar graphs is an entirely different story. Getting a graduate degree barely boosts attendance for women - maybe a percentage point or two. But for men, the boost is huge - about five percentage points. There is no larger gender gap in attendance than between those with a graduate degree. About 29% of men are predicted to be weekly attenders compared to only 26% of women.

Something seems to happen between that one step in education that needs some more investigation. A man with a graduate degree is more religiously active than any other group in this analysis. It does bump up religious attendance for women, too. However the effect is incredibly modest. For women, the bigger jump is between a two year and four year degree.

But let me just take a really big step back here and make this point abundantly clear: education leads to more weekly attendance. For both men and for women. Among people without a high school diploma, about 16% attend weekly. Among those with a graduate degree it’s 29% of men and 26% of women.

Religion has become a luxury good.

Code for this post can be found here.