An updated regional breakdown of the last US election to include Alaska results and the final results for greater New York City and beyond
By Colin Woodard
In December, we shared the regional breakdown of the 2024 U.S. presidential election, showing that despite the existential stakes – democracy really was on the ballot, and it lost – the vote followed the same pattern as the two that preceded it. The election was decided at the margins, with Trump improving slightly over his 2020 performance in every regional culture in the American Nations model.
Much to the irritation of some readers, we’d held out for nearly a month before releasing this analysis so as to be sure most of California’s (notoriously slow) ballot counting was completed. Still, our Dec. 4 post wasn’t based on the final count, and we couldn’t include Alaska’s results because they aren’t reported at the county level. We knew the takeaway would be the same, but that the final numbers might shift margins of victory a few tenths of a percent.
Taking advantage of the summer lull, we revisited the now-final results. To incorporate Alaska’s results, we coded that state’s senate districts by American Nations region, and applied the presidential results in each to the analysis. The final, official results can be seen in the map at the top of the post.
As you can see, the changes are tiny, and apply only to the three regional cultures that extend into Alaska (Far West, First Nation and Left Coast) plus New Netherland because, surprisingly, a large number of New York City area ballots remained uncounted when we ran the data at the end of November. Official tallies for other big cities shifted the final results for Tidewater and Yankeedom a tenth of a point toward Trump and the Midlands a tenth of a point toward Harris. The largest changes overall are for Left Coast (adding an additional 1.2 points to Biden’s margin of victory there); New Netherland (an additional 0.8 for Biden); and Far West (0.6 more for Trump.)
Adding Alaska also gave us data for First Nation, revealing Biden won that indigenous-dominated section of the state by 3.8 points. A cursory analysis of the senate district-level results suggests a racially polarized electorate there, with Biden’s margin going down as the percentage of white people in a district goes up. It’s also likely Biden’s margin would be a bit higher if we were using county-equivalent units instead of (more gerrymandered) state senate districts, as a couple populous districts assigned to First Nation included large swaths of Far West territory.
Using this data, we also created this map allowing for side by side comparisons of the results of the 2024 presidential election in the U.S. and the 2025 federal elections in Canada, subject of this post we published in June. The parties and candidates were different, of course — the Tories are not MAGA and the Liberals aren’t Democrats – but you can see the consistency in regional behavior on both sides of the border. The Midlands are the most competitive region in both countries – the only competitive one, really – while all other regional cultures voted in tandem. (The exception: New France’s “Cajun” enclave in southern Louisiana, once politically distinct, but which has assimilated into Deep South over the past two generations.)
The big difference between the two federations, as we pointed out in June, is in their mix of regional cultures (and population balance therein.) Canada has no Deep South, Tidewater, or Greater Appalachia and, thus, no nationally-competitive authoritarian movement.
Thanks, as always, to our counterparts at Motivf, Tova Perlman (for her electoral data wrangling) and John Liberty (for the maps.)
— Colin Woodard is the director of Nationhood Lab at Salve Regina University’s Pell Center for International Relations and Public Policy.