For as long as pollsters have been ranking groups of Americans, atheists have finished at or near the bottom of every list they make. Least trusted. Coldest rated. Last hired for the jobs that involve other people’s children. And the strangest part is who is doing the ranking: people who, by their own admission, mostly don’t know any of us. They are not rating neighbors. They are rating an idea.

I’ve been writing about that gap for a long time. Back in 2014 I put it this way, in a post called Atheism: Perception is Not Reality: “Time after time, in poll after poll, we find that we are less trusted than – rapists?” And the difference between us and the people doing the judging comes down to one thing, which I said again in 2023: just one: I don’t believe in gods. You do. That’s the whole difference. Everything else on this page is what America has built on top of it.

The rankings

In 2006, University of Minnesota researchers Penny Edgell, Joseph Gerteis and Douglas Hartmann published the study that still frames this subject. They asked Americans which groups shared their vision of American society, and atheists finished dead last — below Muslims, below recent immigrants, below gay people, below everyone. Nearly four in ten, 39.6%, said atheists did not share their vision at all. And 47.6% said they would disapprove if their child married an atheist — the highest rejection of any group on the list. The researchers’ conclusion was blunt: atheists are where America draws the line around who counts as a moral citizen.

Then there is the study behind my rapist headline, and it deserves to be told straight, because it wasn’t a trust poll. In 2011, Will Gervais and colleagues at the University of British Columbia described a man who hit a dog with his car and drove off without helping it, then asked people what that man probably was. Given a list of possibilities, respondents said “atheist” 48% of the time. “Rapist” got 42%. Muslim got 15%, Christian 4%. An untrustworthy act simply looks like an atheist to the average person answering a survey. I wrote about it in 2018 — Why Are Rapists More Trusted Than Atheists? — and the mechanism the researchers identified still seems to me the key to this whole subject: it is distrust, not dislike. The assumption underneath is that people only behave themselves because they think God is watching. Take away the watcher, and in the public mind you’ve taken away the brakes.

The same study had a quieter finding that may say even more. The hypothetical atheist was acceptable as a waitress — 65% had no problem — but as a daycare worker, somebody who watches your children, acceptance fell to 33%. The distrust isn’t spread evenly. It aims precisely at the roles built on trust.

And it doesn’t wash off. When Edgell’s team repeated the work a decade later, on 2014 data, the verdict was “strong, persistent.” A plurality of Americans still looked negatively on the growth of the nonreligious. The wall had been surveyed twice, ten years apart, and it was the same wall.

The thaw that isn’t

Now the better news, such as it is. Pew’s feeling thermometer, which runs from 0 to 100, had atheists at 41 degrees in 2014 — colder than every other group measured. By 2017 we were at 50, and in 2019, 49. Jews, for comparison, sit at 63. So: no longer frozen, still well below room temperature. The coldness runs in both directions, by the way — white evangelical Protestants rated atheists at 25 degrees in 2014, and atheists returned the favor at 28. Neither side is imagining it.

In 2022 Pew changed its method and started asking favorable or unfavorable instead. The most interesting number in that report isn’t about us at all: 55% of Americans said they felt neutral toward atheists, or didn’t know enough to have an opinion. Hostility is giving way to indifference in the middle of the country. That might matter more, in the long run, than any average.

The cleanest long line in all of this research is Gallup’s ballot question: would you vote for a well-qualified atheist for president? In 1958 the answer was yes for 18% of Americans — atheists polled below Black candidates and below women. By 1983 it was 42%. In 2007, 45%. In 2015, 58%. In 2020, 60%. Sixty years of that question describes real movement. But look at what’s underneath it: in 2020, Democrats said yes at 69%, independents at 68% — and Republicans at 41%. Atheists finished second from the bottom of Gallup’s list, with only socialists below us, at 45%. The atheist penalty in American politics is now mostly a one-party penalty.

Which raises the question I can’t leave alone, because it’s my side of the street. Why are there so few conservative atheists? Years ago I picked up a number somewhere — something like ninety percent liberal, ten percent conservative — and I’ve repeated it ever since. So I checked my own memory, the way I’d check anybody else’s. The ten was real: in Pew’s 2014 study, atheists called themselves 56% liberal, 29% moderate, and exactly 10% conservative — 69% Democratic to 15% Republican by party. Today it’s starker, not softer: 82% of atheists lean Democratic, 67% call themselves liberal, and in 2014 atheists made up 1% of Republicans and 1% of conservatives. A rounding error inside the coalition.

Is the conservative share growing? I wanted the answer to be yes; the evidence won’t give it to me. The GOP itself is slowly becoming less religious — its unaffiliated share crept from about 4% in 2006 to around 12% now, mostly through its young — but the unaffiliated have grown more Democratic over the same years, and the only growth in conservative atheists is arithmetic: the atheist population tripled, so a stubborn ten percent is simply a bigger crowd. Education, the usual suspect, is real but unproven as a cause: 48% of atheists hold degrees against 34% of the public, and the label climbs with schooling — yet Pew’s own analysts note that education predicts disbelief, not slack practice, and no study isolates what the degree itself does.

And I don’t fully believe even those numbers. Only about half of Americans who say “I do not believe in God” will call themselves atheists — and the reluctance is steepest on my side of the street: 72% of extremely liberal nonbelievers take the label, just 39% of extremely conservative ones. Republicans who don’t believe in God outnumber Republicans who’ll say “atheist” five to one. Count belief instead of labels and conservatives rise from 10% to about 13% of America’s nonbelievers. The undercount is real, it is worst on the right because the label costs more there — and it narrows the gap without closing it. I’ll have more to say about that closet in a minute. It’s where the numbers and the fear meet.

What it sounds like up close

The studies tell you the shape of it. The mail tells you the sound. A letter to a newspaper editor I wrote about in 2019 put the terms plainly: “you can believe in God any way you want (Baptist, Catholic, Methodist, etc.), but you must believe” — and, “I don’t recall freedom of religion meaning no religion… Get off of our country.” I saved that one under the title it earned: Get Out of Our Country, Atheists!

The politer version is Atheists Should ‘Shut Up and Go Away’ — the demand that we stop objecting to prayers and displays, because this is “our country.” My answer then stands: run the reversal test. Would the same people sit quietly for Muslim-only or Jewish-only prayers at the courthouse? They just want atheists to go away, and they’ve said so in almost those words.

Sometimes it isn’t mail. Sometimes it’s said straight to your face, and years ago I titled a post with the sentence itself: You’re an Atheist? I Hate You. And the misconceptions keep their own catalogue — that we’re Satan worshippers, that atheism is a cult, that nobody could possibly mean it. I collected the current crop just this year, in Atheist: Say it Out Loud. Add the political version — rally rhetoric that lumps atheists in with arsonists and globalists, as I noted in Dehumanization of Atheists — and a reasonable person might wonder whether “atheist” is even a description anymore, or just a convenient device for drawing a crowd.

Here’s the capper. In 2025, Pew asked Americans how much discrimination various groups face. Only 33% said atheists face a lot of it. Fifty-seven percent said religious people do. Muslims registered 74%, Jews 72%, evangelicals 43% — and atheists, the group that finishes last in the trust studies, finished behind almost everybody in the sympathy study too. The public, on the evidence, does not believe we are discriminated against. Back in 2018 I argued the opposite case in Society Needs Atheists — the studies since have not made that argument any weaker. The distrust is measured. The belief that it exists is not.

The closet

In 2022, writing about why some atheists won’t say the word, I put it as plainly as I could: for some, silence is “entirely practical… a loss of social standing and yes, even in the 21st century, a possibility of losing employment or an opportunity as well as promotion.” That wasn’t theory. Let me give it a face — except I can’t give you the face, which is rather the point.

For years, a woman commented on this blog regularly. Friendly, sharp, one of the regulars. At some point she told me, there in the comments, that she and her whole family were atheists. All of them. And nobody outside that house could know, because her husband held local office — city council, county board, something on that order; the exact office doesn’t matter and I wouldn’t print it if it did. An atheist family, living in the open at home and in the closet everywhere else, because one man’s seat depended on it. She isn’t the only one who’s told me some version of that story, here or over on Twitter. She’s just the one I think about.

The research says her family wasn’t being paranoid. In data analyzed by Hammer and colleagues, about 41% of self-identified atheists reported experiencing discrimination in the previous five years — slander, ostracism, coercion, opportunities quietly denied. Studies of concealment find atheists managing their answers audience by audience — one version for family, one for friends, one for coworkers — and the closet runs thickest in the South. I ran a small, unscientific version of the experiment myself, here in Tucson, and wrote it up as Would You do Business With an Atheist Company? Five friends, one question: two air-conditioning companies, same price, one with Christian symbols all over the truck, one plain. Three of the five picked the Christian outfit as more trustworthy. Then the mirror question — an overtly atheist company — and five out of five said they wouldn’t use it. Two of those five had been doing business with me for years. My truck just doesn’t carry stickers.

Now, the honest legal picture, because a skeptic owes you the whole of it. Firing someone for being an atheist is already illegal for most employers — Title VII’s ban on religious discrimination covers non-belief, and the EEOC says so in so many words. Eight state constitutions still carry clauses demanding that officeholders believe in God — Arkansas, Maryland, Mississippi, North Carolina, Pennsylvania, South Carolina, Tennessee, Texas — but they’ve been dead letters since 1961, when a unanimous Supreme Court threw out Maryland’s in Torcaso v. Watkins. I took the harder line myself in Back in the Atheist Closet?, and the fair version is this: the words are still in those constitutions, the force isn’t. But notice what the law can’t reach. No statute fixes the promotion that goes to somebody else, the clients who drift away, the seat that can’t survive the truth. The fear was never mainly about being fired. It’s about everything softer than that.

And so the loop closes, and it’s the saddest mechanism in this whole subject. Pew found in 2019 that simply knowing an atheist is worth thirteen degrees of warmth on that thermometer — one of the largest familiarity effects they measured. Contact works. But the distrust is exactly what keeps atheists from being known. I saw my own small piece of it in 2018, writing It’s Time for Atheists to Change the Perception of the Public: “I think that most of the people I know are unaware of my atheism.” A group can’t be warmed up to by people who don’t know it exists in their own lives — and it can’t safely announce itself to people who’ve already decided it’s the boundary of the moral community. That is the trap, stated in one paragraph.

Tomorrow

So that’s the view from outside: a group that has climbed from 18% to 60% at the ballot box and from 41 degrees to just under 50 on the thermometer — and is still, on the best evidence we have, the last group America trusts and the first it suspects. The perception thawed. The boundary held.

But here’s what the perception studies can’t tell you: while all that distrust was being measured, the group itself was quietly multiplying behind the wall. How many atheists are there, really? How fast did that happen, what made it stall, and how does America stack up against the supposedly godless Europe everybody points to? That’s Part Three, on day three — and some of it is not what you’ve been told. Tomorrow, though, Part Two stays closer to home: the question this post keeps raising, the one I can’t put down. Where are the conservative atheists — and how many of the missing are simply hiding?