Latino voters and the 2026 midterms: what the NBC/Telemundo poll found

Asked which party they preferred to control Congress, 59% chose Democrats and 35% Republicans, a 24-point Democratic advantage. Trump’s job approval among respondents was 27%, compared with 69% disapproval.

The congressional preference result is stronger for Democrats than the poll’s finding at the same point in 2024, when their lead was 12 points. At the same stage in 2018, the Democratic advantage was 43 points. Those figures compare preferences about control of Congress, not presidential approval.

The 2024 presidential vote is a different measure. Exit polls cited by CNN and described in ABC17’s report put Trump’s share of the Latino vote at 46% and Kamala Harris’ at 51%. That result does not directly answer which party voters want to control Congress in 2026.

Nor does the latest poll erase the longer-term picture. NBC News/Telemundo reported that Republicans had made gains among Latino voters, while negative views of the Democratic Party had not improved since 2024. The current Democratic lead and those longer-running party trends can both be true.

What are Latino voters saying about the economy?

The cost of living ranked high among respondents’ concerns. A total of 33% named inflation and living costs as their single most important issue, and 52% placed them among their top two issues.

On household finances, 63% said their family income was falling behind the cost of living. Another 25% said it was keeping pace, while 11% said income was growing faster than costs.

Respondents also gave Trump low marks on the economy. Seventy-eight percent disapproved of his handling of the cost of living, and 74% disapproved of his handling of the economy. Asked how his policies had affected their own economic circumstances, 65% said conditions had worsened for them, 23% said they had improved and 11% said there was no significant difference.

These are separate survey measures: issue priorities, reported household finances and evaluations of Trump’s performance. The poll does not establish whether the people most concerned about prices were also those who said their income was falling behind or who disapproved of Trump’s economic handling. The responses are voters’ assessments, not an independent calculation of household costs or federal policy effects.

Taken together, the figures show why the poll should be read as a set of reported concerns rather than one explanation for the party preference result. It records widespread concern about costs and negative assessments of Trump’s economic performance, but does not establish what caused an individual respondent to favor one party over the other.

What does the poll say about immigration enforcement?

A majority of respondents said the Trump administration’s immigration enforcement had gone too far. Separately, 69% disapproved of Trump’s handling of immigration. On border security, his strongest of the measured policy areas, 39% approved and 58% disapproved.

Views also differed depending on the enforcement question. The survey found that 83% supported deporting undocumented immigrants convicted of serious crimes, while 37% supported deporting all undocumented immigrants. Those responses do not amount to a single yes-or-no position on immigration enforcement.

Some respondents reported concerns that reached into their daily lives. In its account of the NBC News/Telemundo poll, Axios reported that 31% said they or someone close to them had started carrying identification because of enforcement concerns. Twenty-seven percent said they knew someone who had been detained or deported, and 17% said they were considering not voting in person because of enforcement fears.

Those percentages describe what respondents reported about themselves or people they know. They are not verified counts of detentions or deportations, and they do not show how many people ultimately will avoid voting in person.

The poll also asked what respondents intended their midterm vote to signal. Fifty-four percent said it was intended to oppose Trump, 21% said it was meant to support him and 23% said neither. These are stated intentions, not ballots cast or a guarantee of how respondents will vote.

How much can this poll tell us about the midterms?

The NBC News/Telemundo survey measured registered voters, not just people considered likely to vote. Its results cannot establish whether the preferences it recorded will translate into ballots or determine control of Congress.

A separate CBS News/YouGov poll, reported by Axios, found that Hispanic likely voters preferred Democratic House candidates by 60% to 39%. That survey used a different population, so its result is context, not a direct replication of the NBC News/Telemundo poll.

Preferences also differed within the NBC News/Telemundo sample. Women favored Democratic control of Congress over Republican control by 62% to 31%; among men, the split was 53% to 39%. Men were somewhat more approving of Trump than women, but disapproval exceeded approval in both groups.

Generation showed another divide: among third-generation-or-later respondents, 53% preferred Democratic congressional control and 41% Republican control. First- and second-generation respondents favored Democrats by wider margins. The demographic report also said the Florida sample was too small to provide exact state-level figures.

That limitation matters when considering party strategy. NBC News/Telemundo reported that Republicans had built some House strategy around Trump’s 2024 gains among Latino voters and drawn new districts in Florida and Texas with large Latino populations. The report said some districts that had appeared safe based on 2024 results had become competitive; this poll does not provide precise estimates for Florida or individual districts.

The published reports also differ on the survey’s margin of error. The main report lists 3.1 percentage points, while its demographic companion lists 3.5 points. Because the figures conflict, neither can be treated here as a settled single margin.

The unresolved question is whether these preferences among registered Latino voters will show up in the actual midterm electorate. This survey describes opinions; it cannot answer who will cast a ballot or how congressional races will turn out.