I sat down this past week with Carlos Odio, co-founder of Equis Research, who since we met back in 2017 has been the most insightful and honest analyst of the Latino vote I know.
It’s well-known that Trump’s apparent gains with Latinos in 2024 evaporated after he started terrorizing immigrants and wrecking the economy. Right now, Equis has Democrats at about 55% with Latinos on the two-way generic ballot in the competitive House battleground, a rebound from the 49% Democrats got in those same seats in 2024. But it’s not a full recovery: those same districts gave Democrats 57% in 2022, 59% in 2020, and 67% in 2018.
What’s more, those percentages aren’t the full picture. I used the concept of CDM (Contribution to Democratic Margin) to show how much the Latino vote actually contributed to Democratic victories in key states over the last several election cycles. It takes into account not just how wide a margin Latinos supported Democrats by, but also how large a slice of the electorate they made up.
In particular, Democrats will benefit when Latinos who don’t always vote in midterms turn out in higher numbers. Carlos polled Latinos who didn’t vote in 2024 right after the election, and a plurality said they’d have gone for Trump if they had — but those same people are now breaking two-to-one for Democrats in the generic ballot. Historically about 9% of midterm voters overall didn’t vote in the prior presidential, but in 2018’s surge it was 14%. Carlos’s guess for Latinos in 2026 is 12-15%, which could make a difference in key races.
We talked about how “immigration” is actually (at least) two different issues — how to treat immigrants who already live here, and how to treat those who want to come. Those concepts are called “stock” and “flow,” respectively, in academia. And Latinos, like most other people worldwide, are more favorably inclined to stock arguments than flow ones. In 2024, flow (“the border”) dominated the discourse; in 2026, the Trump regime’s monstrous treatment of immigrants and citizens alike will likely be top of mind.
We discussed other nuances about the Latino electorate that most political analysts miss. In particular, Carlos noted that Latinos aren’t a monolith, and how strongly they support Democrats depends greatly on how strongly they identify as Latino. It also depends on things like age (younger is more Democratic) and how far removed they are from the immigrant experience. The Obama-era Latino electorate was more heavily first-generation immigrant, whereas today’s is more second-generation — which doesn’t necessarily mean “more conservative,” but it does mean less predictable.
We also talked about how network effects and the media landscape can change partisan trends among Latinos. Anyone, regardless of their demographics, is more likely to consider voting for Republicans or Trump if they see others like them doing the same. Among Latinos, it’s become more socially permissible to support Republicans for many reasons — including Latinos running for office as Republicans at the local level when they were shut out of establishment Democratic pipelines. And Carlos discussed how the collapse of Univision’s near-monopoly (plus its greater Trump-friendliness) and the rise of ideologically mixed Spanish-language YouTube is reshaping the Latino media diet.
