After the 2024 presidential election, the headlines suggested that Latino voters were moving toward the Republican Party. Donald Trump’s gains were interpreted as evidence of a possible political realignment, particularly among working-class Latino voters and younger men.

Two years later, the story has reversed. Latino voters, we are told, are returning to the Democrats.

Both interpretations risk making the same mistake: treating changes in voting behavior as though they necessarily reveal changes in political identity.

Political scientists have long distinguished relatively enduring partisan attachments from shorter-term influences on elections. Realignment refers to more durable changes in political coalitions. The problem is not the concept itself, but the tendency to interpret changing election results as evidence that such a transformation has occurred.

I spent decades teaching voting behavior at California State University, Bakersfield, at the southern end of the Central Valley. As a public-opinion researcher, I also conducted dozens of surveys throughout the region, including communities that would eventually become part of Republican Congressman David Valadao’s district.

That experience taught me to be cautious about treating demographic characteristics or party registration as reliable predictors of voting behavior. The people who participate in an election are not necessarily representative of the larger population, and their choices can reflect considerations that national partisan statistics overlook.

The Latino Vote Reverses Again

The 2026 midterms produced another substantial change in Hispanic voting patterns.

According to the Associated Press Voter Poll, approximately six in ten Hispanic voters nationally supported Democratic House candidates, while about one-third supported Republicans.

That represented a reversal from 2024, when Hispanic voters were nearly evenly divided between Donald Trump and Kamala Harris.

The longer history makes the problem clearer. Pew Research Center’s validated-voter research found that Hispanic voters supported Joe Biden over Trump by 61% to 36% in 2020. Four years later, Harris received 51% and Trump 48%.

These estimates come from different surveys and, in 2026, a congressional rather than presidential election. They should not be treated as a precise measure of individual vote switching. But the broad pattern is clear: Latino voting margins have changed substantially across recent elections.

If each change represented a lasting transformation in partisan loyalties, Latino voters would appear to have repeatedly remade their political identities in just six years.

Election results cannot establish that.

Some individuals may have reconsidered their candidate preferences. Others may have stayed home, while people who did not participate in the previous election voted in the next one.

A voter can identify as a Democrat and support a Republican presidential candidate. An independent can vote for one party in a presidential election and another in a congressional election.

Party identification, candidate preference and participation are different dimensions of political behavior.

A July 2026 Pew Research Center analysis found that although Trump made substantial gains among Hispanic voters between 2016 and 2024, the partisan composition of Hispanic adults changed relatively little.

A May 2026 UnidosUS survey reported by CBS News found that one in four Latino voters who supported Trump in 2024 said they would not vote for him again. Yet Democrats had not necessarily won their allegiance.

These findings illustrate why changing candidate preferences should not automatically be interpreted as changing partisan identities.

What the Central Valley Taught Me

As I discussed in an earlier examination of California’s 22nd Congressional District, Republican David Valadao repeatedly won elections in a heavily Latino region that also supported Democratic presidential candidates.

In 2016, Hillary Clinton carried his then-congressional district by more than 15 percentage points. Valadao nevertheless won reelection by approximately 13 points. Four years later, he returned to Congress even though Joe Biden carried the district.

Valadao, a Hanford dairy farmer, emphasized agricultural interests, water policy and long-standing community relationships. His victories demonstrated that presidential preferences and congressional voting could produce very different outcomes in the same district.

In 2026, however, Valadao faced a redrawn district and a different national political environment. Democrat Randy Villegas defeated him by a substantial margin, according to preliminary returns.

His earlier victories and subsequent defeat illustrate the interaction of voters, candidates and political circumstances. They do not, by themselves, establish which of those factors mattered most.

What Elections Can—and Cannot—Tell Us

In a March 2026 analysis, political scientists Matt Barreto and Gary Segura raised a familiar question: Were Latino voters changing their preferences, or were differences in turnout producing different electoral outcomes?

Their assessment was that both processes appeared to be occurring.

This is precisely the distinction that matters.

Election returns tell us who won and how voting margins changed. Surveys help us understand attitudes and motivations. Statistical models estimate how demographic groups voted.

But these measures do not necessarily reveal whether the same individuals changed their preferences, whether their partisan attachments changed, or whether a different electorate participated.

A panel study could help answer those questions. By interviewing the same people across elections, measuring their partisan attachments and candidate preferences, and accounting for new participants, researchers could distinguish individual changes from changes in the composition of the electorate. With respondents’ consent, voter records could also help verify participation.

The movement toward Republicans in 2024 was real. So was the movement toward Democrats in 2026. What remains uncertain is how much of either shift reflected changing political attachments, short-term candidate preferences, or differences in who participated.

The mistake is not recognizing that Latino voting patterns changed. It is treating those patterns as an explanation of the voters, rather than something we still need to explain.

–

Kent C. Price is a retired political science professor with four decades of experience in applied public‑opinion research, including extensive work on large two‑thirds revenue elections in California.