The More We Know, the Less We Need Gods

There’s a striking pattern in the data on who believes: the world’s scientists are far less religious than the public, the most educated and secure societies are the least religious, and belief has fallen fastest in exactly the places where information became easiest to get. The pattern is real. It’s also more complicated than either side usually admits, so let’s look at it honestly.

When Knowledge Belonged to the Clergy

For most of history, religious institutions were also the educational ones. In medieval Europe, monasteries held the books and clergy held the literacy; in the Islamic world, learning centered on institutions like Baghdad’s House of Wisdom. Questioning doctrine meant questioning the only educated class there was, and explicit atheism barely existed.

Then knowledge escaped. Gutenberg’s press made books cheap, literacy spread, and Enlightenment philosophers like Hume and Spinoza subjected religious claims to the same scrutiny as everything else. Newton showed the heavens ran on discoverable mathematics. Darwin showed life’s design needed no designer. Compulsory schooling put reading in every hand, and eventually the internet put the whole accumulated argument a search away. At each step, ordinary people depended a little less on authorities to tell them what was true, and in the societies where that process went furthest, belief declined most.

What the Numbers Actually Show

Scientists. Pew’s survey of American scientists found 33% believe in God, against 83% of the general public, and the gap widens at the top: Edward Larson and Larry Witham’s 1998 survey of the elite National Academy of Sciences, published in Nature, found only about 7% believing in a personal God, and a 2013 survey of Royal Society Fellows found similarly low belief among Britain’s most eminent scientists. The people who spend their lives testing claims against evidence overwhelmingly find the God claim unpersuasive.

Societies. The most consistent international finding, documented across decades of World Values Survey data and analyzed by political scientists Pippa Norris and Ronald Inglehart, is that secure, educated, developed societies are markedly less religious than insecure ones. Denmark, Sweden, the Netherlands, the Czech Republic, and Japan sit near the top of both development and non-belief; ten countries and territories now have unaffiliated majorities, up from seven a decade earlier.

Change over time. In the United States, the religiously unaffiliated grew from roughly 5% in the early 1970s to about 16% in 2007 and roughly 28% by the 2020s, before the decline recently slowed and may have leveled off, an honest wrinkle worth knowing. Globally, Pew’s landscape study counted 1.9 billion religiously unaffiliated people in 2020, about 24% of humanity, up from 1.6 billion in 2010, growth driven mostly by people leaving Christianity.

Why Would Knowledge Loosen Belief?

Several mechanisms are easy to see once named. Learning natural causes removes the jobs gods used to do: lightning becomes electricity, disease becomes germs, recovery becomes base rates and placebo effects. Learning statistics reveals that among millions of daily prayers, some will be followed by good news by chance alone. Learning comparative religion shows you thousands of traditions making incompatible claims with equal confidence, which makes it harder to assume yours is the exception. Learning textual history shows the scriptures being assembled and edited by human hands. And learning how our own minds fool us, pattern-seeking, agency detection, confirmation bias, explains why belief feels so compelling from inside even when the evidence is thin.

The Honest Complications

Now the caveats, because the case is stronger with them than without. Correlation isn’t a one-way street: security and education loosen belief, but less religious people also sort themselves toward education, and both directions likely operate. Within a single country the education effect is modest; in American surveys, college-educated Christians actually attend church at similar or higher rates than less educated ones, and the sharp gaps appear between societies and at the scientific elite, not between your neighbors with and without degrees. And globally, the religious share of humanity isn’t collapsing: religious populations have more children, so projections show belief holding roughly steady worldwide even as it fades across the developed West. We covered that demographic story in Why Belief Fades as Knowledge Grows.

What Would We Expect to See?

Here’s the question that survives all the caveats. If belief in God were supported by good evidence, more education and better access to information should strengthen it. The people best trained to evaluate evidence should believe most, and the societies with the freest flow of information should believe most confidently. We observe the opposite, everywhere we can measure it. That doesn’t prove no god exists. But it’s exactly the pattern you’d expect if belief runs on inheritance, community, and comfort rather than on evidence, and it’s very hard to explain if belief runs on evidence. The data doesn’t shout. It just keeps pointing the same way.

Related Arguments

Want to Go Deeper?