The data that made nobody happy
Pew’s 2015 report projected that by 2050, the global share of religiously unaffiliated people will decline from 16 percent to 13 percent. Not because people are deconverting less, but because secular populations have far fewer children. We published the data. Both the secularists and the evangelical revival accused us of being on the other side.
The honest picture is more disorienting than either camp admits. Secularization is real. Religious growth is real. They are happening simultaneously, in different places, driven by different mechanisms, and neither trend is about to reverse.
Three variables
Three variables predict religious trajectory better than any single theory. First, existential security — the Inglehart-Welzel map tracks this with remarkable precision. Second, fertility — religious populations reproduce at higher rates in virtually every context studied. Third, state-religion entanglement — when religion is enforced by authority, backlash secularization follows within a generation. Iran and Ireland are structurally identical cases separated by theology.
The mistake both camps make is treating their continent as the protagonist. The secularists cannot explain why Singapore remains religious despite modernization. The evangelical revival cannot explain why Iran — forty-five years of theocratic education — shows some of the highest private secularization in the Middle East.
The data suggests something none of the four camps has reckoned with: belief, practice, identity, and community function move independently. Americans who stop attending church do not stop believing. Scandinavians who report no belief still baptize their children. Chinese underground Christians display the highest-intensity combination of all four. Treating religiosity
as a single dial going up or down is the simplest version. The real map has at least four dimensions.
Where we concede ground: We have better maps than explanations. Every theory we have is a partial map of territory more complex than our models.
What would change our mind: Sub-Saharan Africa achieving material security without religious decline over thirty years — proving secularization is European, not human.
Read the full synthesis: Where is religiosity growing, where is it receding, and what does that tell us?