Religion in 2024: The Plateau Is Real

But there's no evidence of a religious revival, either.

There’s this very much overused saying in my line of work - data is the new oil. When you hear someone giving a ‘hype’ speech at a data science conference that line usually comes up one way or another. The assumption here is that with the right kind of data, the right kind of analysis, and the right kind of interpretation it’s possible to make decisions that can save lives, lower carbon emissions, avert wars, and make the world a better place.

My aims in life are a whole lot more modest - just describe the wonderful and maddening world of American religion. That’s it, really. But even in my little niche of a subfield of social science, I still need oil. I can’t tell you how many times that I am in the middle of a talk and I’m pointing at a graph, describing a result and thinking, “Man, I wish I could have an even more recent data point to figure out this trend line.”

Well, that day has come my friends. The 2024 Cooperative Election Study has been released. It’s Christmas, New Year’s, Easter, and my birthday — all on a random Tuesday. Here’s why this is a big deal - the CES is the largest longitudinal survey that exists. During an election year, its sample size is at least 60,000. When I combine every single CES data file from way back in 2006 through 2024, the result is a dataset of nearly 700,000 Americans.

You can expect me to strip mine this new 2024 data over the next year or so, but before I get into the weeds I wanted to provide you all with a high-level overview of how much (or how little) has changed in the last year.

Let’s start with the top line finding for me - I continue to double and triple down on a statement that I made about a year ago - the rise of the nones is essentially over for now. Let me show you what I mean.

In 2008, about 21% of the entire sample identified as atheist, agnostic, or claimed no religion in particular. Just six years later the share of the sample who were non-religious was 28%. A jump of seven percentage points in a very short window of time. Six years later, it was up to 34% - up six points from 2014 and a thirteen point increase since 2008. The non-religious share of the population was growing about as fast as anything in the world of demography.

But look back at that bar graph for both 2019 and 2020. The overall share of nones was 35% during that time period. For the samples collected in 2021, 2022, and 2023 the portion who were non-religious was between 35% and 36%. In the most recent data collected less than six months ago the percentage was down to 34%. In other words, there’s been no discernible change in the nones in the United States in the last five years. If anything, there’s a hint that things are actually moving very slowly in the other direction.

The share of Americans who are atheist? It hasn’t really changed at all in the last decade or so. The modal percentage has been 6%. The same is true for the agnostic share, too. It’s not changed in a substantive way since 2012, really. Maybe up a single percent. It’s the nothing in particulars that shifted the most - but their share was just 21% in 2024 - that’s lower than it was in 2019.

But how is this happening? A generational analysis is always helpful.

Let me just start with the Baby Boomers. The share of non-religious Boomers grew slowly and steadily over time from 17% in 2008 to 28% in the 2023 data. But look at what this new survey reveals - the share of Boomers who are non-religious is now just 24%. That’s back to the 2018 levels. And that same general trend is there for Generation X, too. They rose from 25% to 35% between 2008 and 2022. But the estimate in 2023 was 34% and it’s down to 31% in the 2024 results. That’s the same percentage as 2012.

For Millennials, it’s basically the same story as Generation X - it’s just that the overall percentages are higher across the board. There was a ten point increase in the nones between 2008 and 2022, but then a two point decline to 42% in both 2023 and 2024. And Generation Z is a weird one because the same 2022 peak is there at 48%, then a big decline - six points in 2023. (For what it’s worth, I thought that it couldn’t be that large and I said so to numerous groups over the last year). Now, the 2024 number is 46%, which honestly makes perfect sense. I would have guessed the real number was around 45%. So this is confirmation to me that 2023 was just a wonky sample for Gen Z.

Let me put the “I am going to take a picture of your slide with my camera” graph right below. This is the one that sort of encapsulates the entirety of religious change in the United States in a pretty succinct way.

To me, this graph really comports well with how I am thinking about things in American religion right now. The first thing to note is that Christianity has just not been doing well in transferring itself down through the next generation - that’s especially true for Protestants. Among the Silent Generation, 47% were Protestants. It was 5 points lower among Boomers, dropped another six points among Gen X, then declined a whopping nine points among Millennials - just 27% of them are Protestants. Gen Z is even worse at only 22% Protestant - a drop of 25 points from top to bottom.

Catholics, on the other hand, look a whole lot more robust. Among Silents, 22% are Catholic. It’s 20% of Boomers, and 18% of Gen X. It's notable that the Catholic percentage seems to have hit a floor - 16% with both Millennials and Gen Z. But overall, about 70% of the Silent Generation are Christians compared to only 38% of Gen Z. Certainly not great news for the Christians who read this newsletter.

Okay - enough about religious belonging. Let me pivot to the other metric that is incredibly relevant - religious attendance.

If the headline from this post is that religious belonging may be seeing a glimmer of hope, that enthusiasm is dampened significantly when glancing at this graph. The overall decline in attendance at houses of worship has continued at a pretty steady pace since 2008. In that sample, 44% of folks attended less than once a year while 31% were attending religious services on a weekly basis. It’s been all downhill from there.

In the 2024 data, about a third of the sample said that they never attended religious services. That’s a four-point jump since 2020 and a fourteen-point increase from 2008. Overall, 55% of American adults attend a religious service less than once a year now - that’s an eleven point jump in sixteen years.

What buoys this a little bit is the share of Americans who are incredibly religiously active. Again, it was 31% weekly attenders in 2008 and it was still 28% by 2016. Three points in eight years isn’t that bad really. In the next eight years, it also dropped another three points. Today, a quarter of Americans are weekly attenders. Low-attenders bested high-attenders by 13 points in 2008. In the 2024 result it had ballooned to 30 points. There’s no evidence of any type of revival in this graph.

But, if I break this down into generations, there is a bit more nuance that is worth conveying.

Okay - I think we can all admit that the very low attendance category has gained ground among older Americans and younger Americans, too. The never/seldom percentage is up thirteen points among Boomers, twelve points among Gen X, and nine points among Millennials between 2008 and 2024. Meanwhile the high attending share is down four points, six points, and two points respectively. Read that last sentence again, please. There’s not been a huge drop in the share of each generation that is incredibly religious engaged. Instead it’s a lot of movement from the yearly/monthly group down to the seldom/never bucket. The strong core of highly engaged religious people is still there.

Here’s the other piece of this that I think is a pretty interesting result - the religious attendance of Gen Z is actually pretty robust compared to other generations. In 2024, 24% of Gen Z were attending weekly - that was two points higher than Millennials and Gen X. While 53% of Gen Z are attending less than once a year, that’s actually the lowest percentage of any generation in this analysis. Maybe, just maybe Generation Z is more religiously engaged than their parents or grandparents.

So - there you go. It’s a really mixed bag for Christians who are hoping for a revival. This is most certainly not a catastrophic set of data where the nones have just continued to explode in size and religious activity has fallen off a cliff. But there's also not any evidence at all of “A Great Awakening” or anything like that. Religion seems to have hit a bit of stasis in the last five years.

What comes next? I have no idea. But I’m already waiting for another release of CES data.

Code for this post can be found here.