How America Sees Atheists, Part Three: The Numbers

Two days ago, in Part One of this series, I laid out how America sees its atheists — last in trust, coldest on the thermometer, suspected first and warmed to last. Yesterday, in Part Two, I chased the question of where the conservative atheists are hiding. Today we stop asking how the group looks and count it: how many non-believers there actually are, how fast that happened, what made it stall, and how America really stacks up against the godless Europe of the popular imagination. Some of it is not what you’ve been told.

But first, a warning about yardsticks, because nearly every argument about this subject is two people measuring different things. There are three numbers in play, and they are not interchangeable. There is no religion — what you answer when a survey asks your affiliation. There is the unaffiliated — Pew’s basket, which bundles atheists, agnostics, and the “nothing in particular” crowd. And there is the atheist label itself — the much smaller number of people who will actually wear the word. Compare layer to layer, or don’t compare at all.

The long climb

The longest clean line in American religion belongs to the General Social Survey, which has asked the same question since 1972. That year, 5% of Americans said they had no religion. It was still just 8% in 1990. Then the line bends: 17% in 2008, 20% in 2012, roughly 23% in 2018, 28% in 2021. Flat for twenty years, then a fivefold climb in thirty.

I have a personal interest in that baseline year, because I was there for it — seventeen years old in 1972, watching the Jesus Movement fill stadiums and the “Jesus Revolution” make the cover of the news magazines. The 5% baseline wasn’t measured in some neutral, ordinary year. It was measured in the middle of a revival. The climb since then isn’t a fall from a timeless normal; it’s measured against one of the most religiously overheated moments of the century — which makes the flat decades that followed it, and then the sudden bend, more interesting, not less.

Pew’s Landscape Studies tell the same story on their own scale: the unaffiliated went from 16.1% of adults in 2007 to 22.8% in 2014 to 29% in 2023-24, while Christians slid from 78% to 62% over the same span. And the label itself — the smallest, hardest number — tripled: atheists were 1.6% of adults in 2007, 3.1% in 2014, and 5% in the newest study. Thirteen million people, give or take, wearing the least trusted word in American public life.

The institutional shadow of all this is Gallup’s church membership line: 73% of Americans belonged to a church, synagogue or mosque in 1937, about 70% for the six decades after — and 47% in 2020, below half for the first time in Gallup’s history. Gallup pins the fall on the rise of “no religious preference,” which went from 8% of adults at the turn of the century to 21%.

Now the honest wrinkle, because a skeptic owes you the part that doesn’t fit the headline. The escalator has paused. The GSS read 25.2% “no religion” in 2024, down from 2021’s 28% — with real caveats about how the survey was conducted in the COVID years. Pew’s own headline on its newest study is that the Christian decline “has slowed, may have leveled off.” Both things are true: the fifty-year climb, and the recent stall at a high floor. Anybody who shows you only one of those is selling something.

Why does the long direction probably resume anyway? Because of who’s replacing whom. Gallup again: “no religious preference” by generation runs 7% among Americans born before 1946, 13% of boomers, 20% of Gen X, 31% of millennials, 33% of adult Gen Z. Church membership: 66% for the oldest cohort, 36% for millennials. Every generation enters adulthood less affiliated than the one it replaces, and Pew’s switching numbers run about six people leaving Christianity for every one who joins. Escalators pause. Escalators built out of birthdays tend to restart.

Who they are

The atheists, per Pew’s newest portrait: 64% men — down from 68% in 2014, so the gap is narrowing; 75% white, down from 78%; 48% with a bachelor’s degree or more, against 32% of Christians; about seven in ten under fifty; roughly eight in ten leaning Democratic — a subject yesterday’s Part Two took apart in detail. The median Christian in America is now 54 years old. The median unaffiliated American is about 38. Two populations, two different centuries of life experience.

Widen the lens to all the nones and the caricatures soften. In the GSS’s 2012 splits, “no religion” ran 24% among men and 16% among women; 21% among white Americans, 17% among Black Americans, 14% among Latinos. The gaps are real; the monoculture is not. And geography closes the loop with the first two parts of this series: non-belief is thinnest in the South, which is exactly where the workplace research says the atheist closet is thickest. The map of non-belief and the map of its social cost are the same map, turned upside down.

The Europe question

Start where I started in 2018, because the warning hasn’t expired: “non-religious doesn’t directly translate to non-belief.” I said it in a post called Atheists Cannot Catch a Break, and it is the key to the whole comparison. Europe’s giant numbers are no religion. America’s famous number is the atheist label. People set those two side by side and announce a chasm. Let’s do it layer by layer instead.

Layer one, affiliation. In Western Europe, the median country is 24% unaffiliated — the Netherlands at 48%, Norway 43%, Sweden 42%, Belgium 38%, down to 15% in Ireland, Italy and Portugal. The Czech Republic, over in the east, is 72% — Europe’s only majority-none country. Britain went from 31% “no religion” in 1983 to 53% in 2016. And here is the kicker that almost never makes it into the American telling: the United States is at 29% unaffiliated — already above the Western European median. On the affiliation layer, America is not behind anymore. Incidentally, Britain also provides the cautionary exhibit: its 2021 census says 37.2% no religion while the British Social Attitudes survey says 53% — the same country, in the same era, fifteen points apart on question wording alone. Any US–Europe gap smaller than that is inside the noise of how you ask.

Layer two, actual belief — where Europe’s lead is real, and smaller than advertised. The Eurobarometer asks Europeans whether they believe in God, in some spirit or life force, or in neither. “Neither” ran 18% across the EU in 2005 and 20% in 2010: France 40%, the Czech Republic 37%, Sweden 34%, the Netherlands 30%, Britain 25% — against Romania at 1% and Malta at 2%. By identity, about 7% of EU citizens call themselves atheist and 16% non-believer or agnostic. Look inside the Netherlands’ famous 48%: 31 points of it are “nothing in particular,” 14 points are atheist, 3 agnostic. Set Europe’s explicit atheism against America’s 5% label and the honest summary is: genuine, self-declared atheism is two to three times more common in the most secular European countries. A real gap. Also a much smaller one than the headline numbers ever suggest.

How did Europe get there? Largely by switching, within living memory. Pew compared what Europeans were raised as with what they are now: the Dutch unaffiliated went from 22% raised that way to 48% today; Belgium from 12% raised to 38%. A median 60% of today’s Western European nones were raised Christian. And the young are furthest along: among 16-to-29-year-olds in the European Social Survey, the unaffiliated are 91% in the Czech Republic, 80% in Estonia, 75% in Sweden, 70% in Britain, 64% in France — and 17% in Poland, because Europe is not one story either. The profile of the European none will sound familiar after Part One: relatively young, highly educated, disproportionately male — in France, 72% of young men report no religion against 55% of young women. One honest gap in the table: European surveys don’t publish non-belief broken down by race the way American surveys do, so that comparison can’t be run. I’d rather leave the cell empty than fill it with a guess.

And the perception? There is no European Edgell study — nobody needed to fund one. Which is not the same as saying the gut-level prejudice vanished there: the one worldwide experiment on it (Gervais and colleagues, 2017, thirteen countries) found the “an immoral man is probably an atheist” intuition nearly everywhere, secular countries included. Europe’s difference looks less like the intuition dying and more like it losing its teeth — no ballot-box penalty worth measuring, no closet economy built on top of it.

I know what some readers will say at this point, because I’ve heard atheists say it for years: Europeans are smarter than we are — more advanced, they grew out of it, we haven’t. I think that’s crap. There’s no IQ gap in this story. What happened in Europe is that people looked at the church — the state churches, the property, the centuries of money flowing one direction — and concluded it was, in large part, a grift. Americans are simply arriving at the same conclusion later, and more of them arrive every year. It’s a big part of why church attendance in this country hasn’t grown in fifty years — flat at best, declining at worst. People have other claims on a Sunday morning now, even if it’s only the NFL. And what the church offers in exchange is, week after week, a harangue about how horrible you are and how much better you ought to be — delivered to people sitting there thinking, with some justice, I’m doing the best I can. They’re bored. Nothing new has been said to them in years. And underneath it all runs the money: it’s not just the televangelists, it’s every basket that ever passed down a pew, and the small shame arranged for you if yours goes by empty.

It was not always like this, and I don’t say that as a figure of speech — I lived through it, and I remember when the thing was alive. In the early seventies the music stopped sounding like my grandparents’ church and started sounding like us: folk, rock and roll, and Larry Norman above all of it — my favorite then, my favorite now, God rest him, and I don’t even believe in God. There were rallies that filled stadiums with young people: Dallas, June of 1972, Explo ’72 — eighty thousand of them in and around the Cotton Bowl, Billy Graham preaching six times in a week, and a closing music festival people called the Christian Woodstock, with Norman himself on the bill. I was meant to be in that crowd myself — a high school senior living outside Houston, four or five hours away — and I backed out at the end. Two of my friends went, and between them and the national news I heard about little else for weeks. We said the touch of the Holy Spirit had come on a generation, and boom, it bloomed. Whatever that was — and I’ve had fifty years to change my mind about what it was — it is not there today, and nobody can tell you where it went. The excitement is gone, the harangue remains, and the basket still makes its rounds. Atheism isn’t growing in America because Americans finally got smart. It’s growing the way it grew in Europe: slowly, quietly, one person at a time deciding he’d rather be somewhere else on Sunday — and discovering he doesn’t miss it.

What it suggests

Step back across all three parts and one relationship stands out: perception lags demography by decades. While the atheist share of America was tripling, the thermometer gained about eight degrees and the marriage-disapproval line barely moved. Numbers move first. Safety moves second. Trust comes last, if it comes.

Which means the closet — Part One’s subject, and the hidden conservative atheists of Part Two — isn’t a sideshow to the data. It’s the data, at kitchen-table scale: the lag between what the population has become and what it’s safe to say out loud, measured one family at a time.

Europe, read properly, doesn’t show a different species. It shows a sequence. Affiliation leaves first — the “no religion” box gets checked. Belief drains more slowly behind it. The explicit atheist label comes last of all. On that reading, America is mid-sequence, not at the starting line: already at European levels on affiliation, a generation behind on belief and on the word itself. The old framing — secular Europe, religious America — is twenty years out of date on the first layer and still fair on the last.

Years ago, in a post called Society Needs Atheists, I wrote that we’re discriminated against more than any other group. Having now spent three days inside the full record, I’d grade my own sentence this way: the distrust and boundary data put atheists at or near the top of every measured list — while the 2025 perception numbers say the public doesn’t believe we face much discrimination at all. Both can be true. The distance between them is, more or less, the subject of this entire series.

I keep coming back to two numbers that bookend the thing. Thirteen million — the Americans who now wear the label in the open, more than at any time in the country’s history. And one family I know of, atheist to a person, who still can’t say it in their own town because of one man’s seat on a local board. The first number is why the perception will eventually have to move; demography is patient, and familiarity is the one solvent every study agrees on. The second number is why “eventually” is doing a lot of work in that sentence. I was seventeen when the baseline was measured, in a revival year, and I’ve watched the line bend for fifty years since. It bends slowly. It bends.

Leave a Reply