The practical upshot for researchers: the anxiety about Christian nationalism measurement is overblown. The literature built on this measure is on firmer footing than critics have suggested.

Why does any combination work? Because Christian nationalism is a coherent worldview: the US has a Christian character, and Christians should exercise dominion. Any set of items that taps those beliefs will capture the concept — because they’re all measuring the same underlying thing.

Then we ran the most direct test: all 330 possible combinations of 4 items from our list of 11, each predicting Trump vote and immigration attitudes. The result? The estimates are essentially indistinguishable from each other.

We also went beyond the BRS entirely. We added items from PRRI’s index and others we thought should be included. Factor analysis of all 12 items: they all load on one factor, with the same exception (separation of church and state). Different measures, same concept being captured.

The BRS measure has 6 items. One of them — support for separation of church and state — consistently doesn’t load with the others statistically. That one should be dropped. But the other five hang together strongly and consistently, across very different datasets and data collection modes.

Reviewers keep pushing back on Christian nationalism research with measurement critiques — one memorably called the standard Baylor Religion Survey measure “garbage.” We decided to actually test whether that criticism holds up. Across 10 datasets from 2007 to 2024, the answer is mostly: no.

I would love to check this out, but I don’t think I’d fit in if I went. Thanks for the shout out!

Sure! Data collection was provided by Cint in Sept/Oct 2025. N~3300. We used weights to bring the sample in line with 2024 census estimates. Happy to answer other questions!