What Does Religion Look Like At Elite Universities?

Is Yale full of secular students, while Midwest regionals are teeming with Christians?

I think people vastly overestimate the unbelievable inequalities in higher education in the United States today. I would hazard a guess that if someone asked a human being on planet earth to name the best college in the world, a large share would say Harvard. Here’s what may be contributing to it’s vaunted status: Harvard is sitting on an ocean of money. How much? In fiscal year 2024, it was $53.2 billion (PDF). Their endowment has grown by at least $10 billion since 2021. The Wall Street Journal described Harvard as “a hedge fund that has a university,” back in 2017.

What really grinds my gears about Harvard (and others in that stratosphere) is that rich people still write checks to their endowment. Malcolm Gladwell did a podcast episode on this phenomenon and titled it, “My Little Hundred Million.” If you write a nine figure check to Harvard, they just throw it on the pile. At a place like Eastern Illinois University, it would absolutely change everything. Our endowment sits at $5.5 million.

The student body at Harvard is not at all like the composition of my classroom. Harvard’s legacy admissions rate is 33%, compared to 6% for a non-legacy. In contrast, at EIU, our motto is, “y’all come.” I don’t think I’ve ever heard anyone describe themselves as an EIU legacy. My guess is that Harvard’s share of first generation college students is a bit lower than my institution. But, I wanted to push beyond simple socioeconomic status. What about the religious composition of elite universities versus regional comprehensives.

Using FIRE’s (Foundation for Individual Rights and Expression), recent survey of a bunch of college and university students - let me show you the religious composition of Harvard and Yale, compared to Southern Illinois University - Edwardsville, which is the epitome of a directional university in a flyover state.¹

There are some pretty big differences when looking at this data. The first is that Harvard/Yale have a whole lot fewer Protestants - just 25% of the student body, compared to 38% at SIU-E. So, that’s about thirteen points that Harvard/Yale need to find somewhere else. A big chunk is in the difference in Catholic students - you are twice as likely to encounter one at these IVY league schools than SIU-E (18% vs 9%). The rest of this is due to there being a whole lot more Jews, Hindus, and Muslims at Harvard. Which does make sense from the perspective of regions. There are larger concentrations of Catholics in the Northeast, and huge pockets of folks who come from non-Christian backgrounds.

However, what really struck me was the non-religious share. At Harvard and Yale, the percentage of the student body who identified as atheist or agnostic was 28%. It was the exact same share of students who were attending Southern Illinois University - Edwardsville. And, the percentage who claimed no religion in particular was almost exactly the same in both (17% vs 15%). It’s pretty noteworthy that there aren’t more nones at elite Ivy league institutions. But is this just a quirk of choosing these three institutions to make this comparison?

Using this database, I chose schools that had an acceptance rate below 30%.² Then I merged that with the FIRE data. Of the 257 schools that were polled in the survey, 54 of them were deemed as selective based on this criteria. Let me show you the religious breakdown of selective vs non-selective schools.

I think the overwhelming sense I get from this graph is how small the differences are between these two types of institutions. The Protestant and Catholic shares are almost exactly the same. That’s also true among Latter-day Saints, Orthodox Christians, Muslims, and Buddhists, too. There are only two groups that significantly diverge. One is the share of college students who just identify as “Christian” on surveys. I’ve written about this before - loads of young Protestants don’t know what that word means. They know that they are Christians and that they aren’t Catholic, either. In this data, 21% of non-selective college students are Christians compared to just 11% at selective universities. The other divergence is Jews - they make up 7% of students at elite schools compared to 3% at other institutions.

When it comes to non-religion, there are differences in the percentages but they aren’t that large. At elite schools, 14% of students say that they are atheists and another 17% indicate that they are agnostics. At non-elite schools, those numbers are 11% and 14% respectively. That means that the atheist/agnostic share is about 6 points higher at prestigious schools. However, the ‘nothing in particular’ percentage is exactly the same. So, the nones are more prominent at places like Harvard and Wellesley, but not by that much.

I was thinking about this a bit, though and had an idea - maybe the differences in religious composition can really be traced back to politics. Liberals tend to be less religious than conservatives and schools like SIU-E tend to be a bit more moderate than places like Vassar. So, if I compared liberal students to other liberal students, would that make the differences in religious composition disappear?

I don’t know if this makes a huge difference to the overall takeaways from the prior graph, but there are a few things worth talking about. Protestants are way overrepresented in non-selective universities, even when holding political ideology constant. For instance, among conservative students at non-selective universities - 45% of them are Protestant. That’s sixteen points higher than the same type of student at elite schools. The Catholic numbers are the other direction, though. It’s 34% of conservatives at selective schools compared to just 23% at non-selective ones. That Protestant gap is also there for those students who are middle of the road (9 points) and liberal (4 points), but it’s quite a bit smaller.

But that’s really the only big divergence that jumps out. Once you control for political ideology, the share who are non-religious at each type of institution is almost exactly the same. Yet, there’s one thing that may be easy to overlook in this graph that is worth flagging. In this data almost 60% of liberal students identify as atheist, agnostic or nothing in particular. That’s true at all types of institutions. For those who are moderate, it’s 35%. For conservative college students, it’s 20%. A liberal college student is three times more likely to be non-religious than a conservative one. Let that roll around your brain for a minute.

Let me shift the analysis to religious attendance now - the FIRE survey offers a bunch of options ranging from never to more than once a week.

Yeah, again, I am not bowled over by any huge differences in the religious attendance of students at Ivy league schools versus non-selective institutions. 49% of students who attend prestigious schools attend church less than once a year compared to 46% of students who go to a non-selective school. So, those at the top end are slightly less religiously active, but three points is certainly not a chasm. That's the general trend here when comparing across all types of attendance levels. For students at non-selective schools, 19% say they attend religious services about weekly or more. It’s 14% of those at selective schools. Again, a gap, but a relatively small one.

But ignore the comparison between these two groups and just focus on the incredibly low levels of religious attendance among college students. About 60% of them attend a house of worship no more than twice a year. About one in ten are there at least once a week. College students are incredibly irreligious.

But let me control for political ideology again like I did in the analysis about religious tradition.

It’s staggering to me how little differences there are when we compare liberals to liberals and conservatives to conservatives. The proportion of liberals who never attend deviates by three percentage points between non-selective and selective institutions. The share who attend weekly is exactly the same. In fact, when you look at moderate college students, about 18% are weekly attendees, whether they attend an Ivy league school or a regional institution. Among conservatives, 33% are at a house of worship once a week. Again, no difference between the two types of institutions.

The huge chasm, again, is between liberals and conservatives. A liberal college student is thirty points more likely to be a never attender than a conservative one. Among conservative college students 32% attend weekly compared to 8% of liberal college students. That’s how big the God Gap is - about 400% on college campuses.

If I were given the challenge of guessing the religious persuasion of a random college student but I couldn’t ask questions directly about religion, I know what I would ask, “What’s your political ideology?” While there are certainly irreligious conservatives and religious liberals, they are certainly hard to find, based on this data.

What I’ve come to conclude is that the primary divide is not between elite and non-elite institutions. It’s really between the political right and political left when it comes to understanding the religiosity of college students in the United States today.

Code for this post can be found here.

1. Absolutely zero shade to SIU-E, by the way. I teach at a school that is in the exact same situation, both in terms of geography and educational prestige. But I believe that schools like ours are some of the strongest engines of economic growth in the United States. We help young people that come from pretty tough SES backgrounds and move them up a rung or two on the economic ladder. I don’t know if Harvard or Yale are doing that in large numbers.

2. Very few universities have this low of an acceptance rate, FWIW. Pew analyzed admissions data from 1,364 universities. They found that just 6% of all universities admitted less than 30% of applicants. A majority of colleges admit most students who apply.