The Rise of Singleness and How Religion Is Impacted

Marriage and Religion are Declining Institutions

A few months ago, I wrote a post about falling marriage rates and the possible link to religion:

But when I read over those graphs I think I had a big blind spot - gender. Obviously marriage rates aren’t the same for men and women for a wide variety of reasons. For instance, an article in the NYTimes in late July was focused on how online communities have sprouted up to help other women know that specific men they find on dating apps are safe to be around.

Dan Cox, from the American Enterprise Institute did some polling and found that younger women (18-29) were significantly more likely to report that they were single in 2022 compared to 2020 (45% vs 38%). And a book published in 2015 called, “Date-onomics: How Dating Became a Lopsided Numbers Game,” found that educated women just didn’t have that many options when it came to finding a potential mate because the share of men going to college has declined sharply in recent decades.

So, I wanted to explore that gender gap on marriage a bit. But also I wanted to see how all of that related back to religion. I think it goes without saying that lots of people have found their current spouse at a house of worship. But is being single driving women further away from religion than unmarried men? These are questions worth some analysis and reflection.

Let’s start with the broadest question: are women more likely to report that they have never been married compared to men? In 2008, about 20% of all women in the sample reported that they had never been married - it was 30% of all men. That’s not a small gap and it’s persisted for the entire length of the Cooperative Election Study. Both trend lines have slowly edged up every year.

However, I would be remiss to point out that the line for women has stayed relatively stable beginning in about 2018 when 26-27% say that they were never married. For men, the number continues to climb. In the most recent data collected from 2022, about 37% of men say that they have never been married. The overall conclusion is pretty unmistakable: singleness is on the rise for both men and women, but women are still 8-10 points less likely to never be married.

This is clearly a function of age, of course. Older people have just had more opportunities to get married compared to younger ones. So I tracked the share who have never been married for both men and women in several years between 2008 and 2022. The huge shift in marriage is apparent in just the last fourteen years.

I think it’s really instructive to look at forty year old respondents. I think that this is an age where if someone is going to be married, they have probably already walked down the aisle. In 2008, among men, about 20% have never been married. For forty year old men in 2022, 37% had never been married. That’s insane to consider — a forty year old man is nearly twice as likely to never be married in 2022 compared to 2008.

For women, there are also huge differences, as well. In the 2008 data, about 12% of women had never been married by their 40th birthday. In 2022, that share had increase to 26%. In other words, a forty year old woman was twice as likely to be unmarried in 2022 vs 2008. That’s just a fourteen year window of time, not decades.

Let’s bring religion into the mix now, though. It certainly is looming in the background of these conversations. Every Western religious tradition has a strong ethic of marriage and family and there’s certainly reason to believe that being single puts one out of step with the predominant culture of religious communities in the United States.

I restricted the sample to just folks who reported never being married and then calculated the share who reported never/seldom attending religious services in the left panel and those who were attending weekly or more in the right panel.

Note that single women are less likely to be never/seldom attenders compared to single men, but both shares are clearly rising. In 2008, about 47% of unmarried women were seldom/never attenders compared to 55% of men. As time has passed both numbers have increased and the gender gap has narrowed significantly. In 2022, 63% of unmarried men were attending less than once a year compared to 60% of women. The gap has gone from eight points to three points. That’s something to keep an eye on.

Also, when it comes to weekly attenders among never married folks, the gender gap is narrowing, too. In 2012, about 25% of unmarried women were attending religious services every week compared to only 20% of men. But in the most recent data that gap is now just 1-2 percentage points. It’s pretty fair to say that single women are catching up to single men on metrics related to religious attendance.

But have there been religious composition changes among never married men and women over the last several years? I explored that in the graph below and was pretty surprised at what I discovered.

In 2010, 57% of never married men were Protestant/Catholic. It was 59% of never married women. The share who were nones was basically the same, too. As time has passed the Protestant/Catholic share has predictably dropped for men and women. In 2022, just 40% of unmarried men were Protestant or Catholic - it was 45% of women. The share of unmarried men who were nones (atheists/agnostic/nothing in particular) has now risen to 52% - for women it’s just a bit lower at 49%. My takeaway is that about half of never married folks are nones now, while Protestant + Catholic is in the low forties in terms of percentage.

Just to point a finer point on this, I calculated the religious composition of people who had never been married vs currently married. I did that analysis for both men and women.

Sixty-two percent of married men and 63% of married women are Protestant or Catholic. Among the never married folks it’s 40% and 45% respectively. Just 31% of currently married men are nones and 30% of married women. Recall that about half of never married men and women claim no religious affiliation. What stood out to me here was that there wasn’t a big difference between men and women in terms of religious composition. Married folks are way more likely to be religious compared to never married folks, gender just doesn’t matter that much.

Let’s get to a regression now to really isolate the factors that lead people to claim no religious affiliation. That’s the dependent variable here - a dichotomous variable of whether someone was atheist/agnostic/nothing in particular or not. I threw some basic controls into this model: education, age, income, race, and partisanship on top of being never married. I ran the model for both men and women.

Interpreting this coefficient plot is simple. Any point estimate to the right of zero means the variable predicts more of the dependent variable (in this case likelihood of being a none). If it’s to the left of zero, it predicts a lower likelihood. If the estimate or error bars overlap with zero, it’s not statistically significant and can be ignored.

This is a really surprising result! For men, being unmarried is strongly predictive of being an atheist/agnostic/nothing in particular. The only variable that is more predictive in the positive direction is being white. For women, there is no statistical relationship between never being married and being a religious none when controlling for a host of other demographic factors. I find that to be fascinating. Singleness is not a gateway to a secular life for women, it seems to be strongly predictive for men.

But, I also ran the same analysis but switched out the dependent variable to never/seldom attending religious services. Again, the interpretation is the same - positive values indicate being more likely to be a never/seldom attender. Negative values are the opposite.

For people who report that they have never been married, there is an increased likelihood of being a never/seldom attender. That’s true for both men and women. However, the relationship is much stronger for men than women. For men, being never married is about five times more predictive than the same variable for women.

Thus, there’s some real evidence here that never being married is intimately related to religion, especially for men. Among men who have never walked down the aisle, there’s data that indicates that they are more likely to be non-religious and more likely to be never or seldom attenders. For women, it’s only the attendance that is affected.

This result seems to be building a larger case that men are just dropping out of many institutions that are key to a modern functioning society. They are significantly more likely to never get married, they are also significantly more likely to be a never/seldom attender and to identify as atheist/agnostic/nothing in particular when they have never been married.

In my mind, I see a lot of these types of institutions that were the bedrock of American society like interlocking pieces. They fit neatly into the other ones. Like having a college degree is strongly linked with religious attendance. Having a middle class income is also linked to increase religious attendance.

Being religious is correlated with having more children. As I found here:

The most stunning result for me is simply this: just forty percent of atheists in their early 40s report having a child under the age of 18. It’s over 70% of Latter-day Saints.

Deciding to deviate from the “proper” path in one way seems to be strongly connected to dropping out of other parts of society. And this result seems to add some nuance - it’s men who are the ones dropping out with more regularity.

I am not one of these folks who wants to write a “woe to men” post. Those are legion on the internet. But I do think it would be wise to consider what society looks like when about half the population is refusing to engage in many aspects of a functioning democracy.

Code for this post can be found here.