Did immigration or big Catholic families keep the US Catholic share steady?

GSS counterfactual: immigrants added +1.07 points [0.52, 1.68] to the 1990-2018 change in the US Catholic share; Catholic mothers' higher fertility added -0.37.

Between the 1990s and 2010-18 the share of American adults who call themselves Catholic barely moved in the General Social Survey: 26.4% to 25.0%. It held while a quarter to a third of people raised Catholic were leaving (retention fell from 78% to 67%). Two stories are usually offered for what held it up: immigrants, mostly from Latin America, and Catholic families having more children. I tested them against each other today.

How. Two counterfactuals from the same survey. For fertility, I asked how many US-born adults would have been raised Catholic if Catholic mothers had had the same number of children as other mothers. The ratio comes from GSS women aged 45 and over: Catholic mothers born between about 1925 and 1960 had about 1.1 times the children of other mothers. For immigration, I dropped the foreign-born. I wrote the prediction down and committed the script before running it: fertility would add about a point to the share in both decades, but a shrinking point, and immigration would carry the change.

What came out.

  • Fertility added 1.17 points to the Catholic share in the 1990s and 0.81 in 2010-18. Immigrants added 2.30 and then 3.37.
  • So over the change from the 1990s to 2010-18, immigration contributed +1.07 points (95% interval +0.52 to +1.68) and fertility -0.37 (-0.69 to -0.02). The difference, +1.44 [+0.74, +2.10], survives every alternative I tried: a different generation length, counting fathers, using the mother's upbringing instead of her current religion, and US-born mothers only.
  • The model has a gap. It explains 2 to 4 points fewer Catholic upbringings than happened for people born 1925-69. Even if all of that gap is credited to fertility, the fertility boost is a flat 2.2 points in every decade, so its share of the change stays at zero.
  • Since 2018 the picture reverses: the immigrant contribution fell by 1.38 points, the largest part of the recent drop to 22.2%. That last step has the web-interview caveat from earlier posts. Correction, 2026-11-13: the caveat turned out to matter. With only the 2022 and 2024 in-person interviews, the fall is -0.44 points (-1.38 to +0.49), not distinguishable from zero; see the follow-up.

Limits. Dropping immigrants is a composition contrast, not a model of a country without them. The fertility side assumes children are raised in their mother's religion and measures her religion at interview, not when she was raising them.

Code, tables and log: research/catholic-demography/experiments/m5h_fertility/. Live page: https://marisstella.app/research/catholic-demography/

All writing · More in Research findings · More on catholic-demography · All tools and research