(ANALYSIS) I have to admit that I was pretty gobsmacked when I saw a post on X about the result of the Harvard Crimson’s annual poll of the school’s incoming class that will presumably graduate in 2027. The graph that grabbed all the traffic was about the political persuasion of these 18 year olds at one of the most elite universities in the world.

Twenty percent identified as very progressive, and another 45% said that that they were progressive. A quarter said that they were moderate, and 8% self-identified as conservative — 65% progressive, 8% conservative. You can see it here. But God bless those crazy kids at Harvard because they also asked the incoming class about religion. Yeah, that’s right in my wheelhouse.

Here’s how they visualized the religious composition of the Harvard class of 2027:

There’s a lot of stuff here that is worth some exploration. The most likely response option chosen by Harvard first-year students? Agnostic. The second most likely? Atheist. In total, nearly half of Harvard’s first-years chose one of those two options. That’s insanely high. For comparison’s sake, about 12% of the general population identifies as atheist/agnostic. So, Harvard is four times higher than that.

But that’s not the only thing that jumped out to me. Maybe the biggest surprise was that just 6% of all Harvard first-years identified as Protestant. That’s insane. In the Pew data from a general population sample in 2022, 43% identified as Protestants. The average America is seven times more likely to be a Protestant than a Harvard first-year.

Here’s what I wanted to do with this post: contextualize the Harvard class just a bit more. How much do they differ from college first-years at other universities? How much do they differ from all 18 year olds? Then, I wanted to show just how much of an outlier Harvard is when it comes to religiosity compared to other universities in the United States.

Let’s start by comparing those numbers from the Harvard Crimson to those collected by the FIRE (Foundation for Individual Rights and Expression) survey of college students that was conducted earlier this year. The total sample was over 55,000. If I restricted it to just 18 year olds, I am working with 3,989 students. That’s a pretty solid N size.

I first have to say that this is not a direct comparison because for reasons that aren’t clear to me, the Crimson uses response options that don’t map exactly on to how almost every other survey ask this question. The biggest discrepancy is that the Harvard poll doesn’t offer the “nothing in particular” option. That’s a big omission, and it’s absolutely going to skew some results. I can’t really say how that works in practice, but it’s worth considering when looking at these graphs.

So, a Harvard first-year student is about twice as likely to identify as atheist or agnostic compared to other first-years at the 250+ universities polled by FIRE. The Buddhist number is way out of spec, too. Buddhists make up 1% of the entire population, and apparently 1% of the first-year student population, too. At Harvard, they are 12%. That’s clearly an outlier.

The aforementioned Protestant figure is just astonishing when looked at through this angle. Thirty percent of college first-years are Protestants — that’s five times the rate found at Harvard. Think of it this way:

At Harvard you are twice as likely to encounter a Buddhist first-year as you are to meet a Protestant one. At any other university in the FIRE sample, you are 30 times more likely to randomly select a Protestant from the first-year class compared to a Buddhist.

The other thing that is worth noting is that this same weirdness is not there when looking at Catholics. They are slightly less prevalent at Harvard as other universities, but you probably wouldn’t notice that difference in a real world situation.

How about I compare Harvard first-year students to all 18 year olds in the 2022 Cooperative Election Study? That will give us a different reference case.

Here, the numbers shift just a bit, but not in a dramatic way. The thing that you should take note of is that in the general population sample of 18 year olds, 34% say that they are nothing in particular. In the FIRE sample of college students, it was only 15%. I’ve written about this a bunch — college is not some sort of catalyst for leaving religion. That whole line of thinking just needs to stop.

The Protestant disparity is a bit smaller here because there are fewer Protestants in the general population sample of 18 year olds (22%). But those huge gaps on Buddhists and Hindus are still just as pronounced. Seventeen percent of Harvard’s first-years are in those two categories. It’s less than 2% of all 18 year olds.

Okay, so we’ve established that Harvard is clearly an outlier when it comes to students’ overall religious composition. But I wanted to see if there are any other schools that come close to their religious make up. The first metric I used was the share of each school that identified as Protestant. I didn’t limit this sample by age, by the way. If I did, I wouldn’t have a sufficient N to do this analysis. So this is the overall composition of the entire student body — not just incoming students.

I have graphed the 25 schools with the smallest share of Protestants out of the 258 universities in the dataset. It should surprise no one that BYU Provo leads on this metric. Just 1% of its student body is Protestant. I would love to meet that person, by the way. I’m sure that they’ve had some interesting stories to tell about their time with the members of the Church of Jesus Christ of Latter-day Saints.

Recall that Harvard’s overall Protestant share was 6% — that’s lower than any other school outside of the aforementioned BYU. There are some who are close like Utah State and UC Berkeley at 8% and 9%, respectively. There are a whole bunch who are around 10-12%, though. They are all over the country, as well — University of Vermont, University of Denver, Boston University. I think it’s fair to say that there a lot of liberal arts schools in New England on this list, but that’s not the majority.

I know some of you are going to ask about this, so here are the schools with the highest share of Protestants.

Liberty University leads the charge at 84% Protestant. Again, meeting the non-Protestants studying in Lynchburg would be interesting. But then there’s a big drop off. It’s 63% at Auburn, 59% at Arkansas State and 58% at Ole Miss. But then there are a bunch of schools right around the 50% mark. I think it’s fair to say that these schools are concentrated in the American South. Arkansas, Alabama, Kentucky, Tennessee, Texas and the Carolinas are all on this list. Not a lot of places out west or in the northeast.

For those wondering, the overall mean for the FIRE sample was 27%. When I restricted the sample to just 18 year olds, it was 28%. So, Harvard being 6% Protestant is undoubtedly low, 4.5 times lower than the overall average.

The other outlier was that atheist/agnostic share. Recall that 46% of all Harvard first-year students identified as atheist or agnostic. Are there other schools in the United States that are at that same level? One thing I want to point out here is that the FIRE survey had the “nothing in particular” option — which didn’t appear in the Harvard Crimson’s survey of their first-years. Functionally, I’m sure this drives up the atheist/agnostic share in Harvard’s data because a lot of nones would just say that they are atheist or agnostic because they didn’t have the “none of the above” option.

There are schools in the sample that are certainly in the same ballpark as Harvard’s student body. For instance, Harvey Mudd’s students are 43% atheist/agnostic. It’s 40% at Bates College and 39% at Worcester Polytechnic Institute. But there are a bunch of schools that are in the 30% range. It doesn’t take a long look at this list to notice a pattern. It’s largely dominated by those very prestigious liberal arts colleges in New England like Colby, Wellesley, Vassar, and Bowdoin. I think it’s fair to say that this category of American higher education is the most secular.

Just for symmetry’s sake, I visualized the universities with the lowest percentage of atheists and agnostics.

There was not a single atheist or agnostic at Liberty University. I know, that’s shocking to the readers of this Substack. But what is more interesting to me is that 5% of students at the University of Mississippi — a large state school with a diverse population — identified as atheist or agnostic. That’s in the same general range as an evangelical school like Pepperdine and not that much higher than Hillsdale. Auburn is in a similar class as Ole Miss, with only 8% atheist/agnostic.

There’s really not a lot of rhyme or reason to this list. There are some southern schools, but that’s certainly not the majority. For instance, this graph includes San Diego State, SUNY Albany and Howard — not a lot of overlapping characteristics between those universities.

In the entire sample collected by FIRE, about 19% of students identified as atheist/agnostic. That number didn’t substantively shift when I restricted it to just 18 year olds. It was 46% of first-year students at Harvard. So, atheists/agnostics at Harvard are 2.5 times higher than at the average university. Again, there is some survey methodology issues at play here, but I still think it’s fair to say that Harvard’s campus is a significant outlier when it comes to the number of atheists and agnostics in the classrooms.

Look, I know that Harvard’s incoming class is not just a random slice of graduating seniors across the United States. That’s not how these things work. But at the same time, I think Harvard should focus it’s recruiting efforts on diversity across a number of dimensions. When I just wrote the word “diversity,” I’m sure many of you automatically thought about racial diversity. And that’s certainly part of the equation.

However, what about religious diversity? When Harvard’s campus is 22% Protestant and Catholic and 46% atheist and agnostic, that’s so far afield from diversity that it’s worth some serious reflection. Of course Harvard has no obligation to seek out evangelicals to join their student body or practicing Catholics. But it’s nonsense to think that the classroom would not be more conducive to learning when a variety of religious opinions are represented.

Think about all the fields that intersect with religious questions — philosophy, law, psychology, sociology, political science, art history, anthropology — and the list can go. Diversity means more than just seeking out non-White, non rich kids. It means trying to reflect what the entire student population of the country looks like. Harvard is clearly missing the mark in this regard. I’m not hopeful that they will earnestly seek to provide a diversity of religious viewpoints in their classrooms or dorm rooms in the near future.

This piece is republished from Graphs About Religion on Substack.

Editor’s note: Apparently the Harvard Crimson made an error and the share of Buddhists went from 12.1% to 1.7% but no other percentage changed. Which means that ~10% of their sample is missing from the analysis. I’m not rewriting this post, because everything beside the Buddhist number stayed the same.

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.