Method
Two public files, one division, and some caveats that matter more than the tables.
The data
- Names: the Social Security Administration's "Popular Baby Names by State" files for 2025, which list every name given to five or more babies of one sex in a state that year. Names given to fewer than five are not in the file, so rare names are undercounted and very small states show fewer names.
- Votes: the Federal Election Commission's official 2024 presidential general election results by state. A state is "Democratic-won" if Harris received more votes there than the other major candidate. Nebraska and Maine split electoral votes by district; this site uses the statewide popular vote.
The calculation
For each name and sex, add up its births across the 31 Republican-won states and across the 20 Democratic-won states (DC counted as a state). Divide each by that group's total named births to get a share. The ratio of the two shares is the figure in the tables. Only names with at least 1,000 recorded births nationally are ranked, so a handful of births can't produce a dramatic ratio.
What it does not mean
- Where a baby is born says nothing about how its parents voted. Every state has millions of voters on each side.
- Many of the strongest leans track region, language and religion, not party: names common in Spanish-speaking, Jewish or Muslim families cluster in some states; names from country music or Western culture cluster in others. The vote is one lens on geography, not a cause.
- Shares, not counts, drive the ratio. California alone records more births than many Republican-won states together, so a name can be more common per birth in one group while having more total births in the other.
Reproduce it
Download the SSA state file (namesbystate.zip) and the FEC spreadsheet (2024presgeresults.xlsx), keep the 2025 rows, group by state winner, divide shares. The code behind this site does exactly that and nothing more. Its twin, republican.baby, runs the same code on the other group of states.