The Economy
Rises to the Top In September 2025, 26 percent of Latino registered voters rated national economic conditions as excellent or good. By the fall of 2026, that share had fallen to 19 percent. Seven points, in one direction, on a question that asks about the national economy rather than about a family budget. The same survey that recorded the lower overall rating also recorded rising concern about specific costs and a larger share blaming a named policy author for the direction of the economy. Both readings describe a minority: about one voter in four a year ago, closer to one in five now.
The new reading comes from a survey of 5,062 U.S. adults, conducted from Sept. 28 to Oct. 4, 2026, in English and Spanish. The sample is built to represent the views of the full U.S. adult population. Of those 5,062 adults, 774 are Latino registered voters. That subgroup is the source of every figure described here, so each percentage below describes the 774 and not the full 5,062. A subgroup of that size carries more uncertainty than a full sample, and that should be read into every number that follows. Everyone who answered belongs to the same standing panel of respondents, the American Trends Panel, which is asked questions repeatedly over time rather than recruited fresh for each survey.
A respondent counts as a registered voter only if the person says they are absolutely certain they are registered at their current address. Anyone who says they are not registered, or who is unsure, is counted as not registered. That standard rests on what the respondent says about themselves. “Other registered voters” means everyone who is not Latino. “Hispanic” and “Latino” are used interchangeably. Validated voters are adult citizens who told the survey they voted in a general election and whose record of voting appears in a commercial voter file (a private database of voter registration and turnout records). In that framework, Trump voters are validated voters who said they voted for Donald Trump in 2024, and Harris voters are validated voters who said they voted for Kamala Harris. Nonvoters are people who did not vote in 2024 but were eligible to do so. Those categories draw the boundary around every percentage that follows, and they decide who is counted inside each one.
Here is the finding that organizes the rest of the picture. A deepening gloom about the economy and a first-place ranking for the economy as an election issue come from the same respondents. Eighty-six percent of Latino registered voters say the economy is very important to their vote for Congress in November. [1] On that measure, Latino voters look like the country as a whole: the economy ranks first for all registered voters, not only for Latino ones. The priority also crosses the party line inside the Latino electorate. Eighty-nine percent of Latino voters who support their local area’s Democratic candidate for the U.S. House of Representatives in the 2026 midterm elections call the economy very important to their vote. Seventy-eight percent of those backing their district’s Republican candidate say the same. For each group, the economy is the top issue, so the ranking is shared rather than split. What separates them is the verdict: 90 percent of Latino Democratic voters rate the economy negatively, against 58 percent of Latino Republican voters who call it only fair or poor.
Below the top of the list, agreement holds for a while and then breaks. At least three-quarters of Latino voters say healthcare, at 80 percent, and education, at 75 percent, are very important to their congressional vote. Both figures sit close to the economy’s 86 percent. Both also behave very differently from it. On healthcare and education, the two coalitions do not rank together. Latino voters who back their Democratic candidate are 31 percentage points more likely than Republican backers to call healthcare very important to their vote. On education, the same gap is 27 points. Set against them, the 11-point spread on the economy — 89 percent to 78 percent — is narrow. Two-thirds or more of Latino voters also name the future of democracy as a key consideration, at 69 percent. Immigration follows at 64 percent. Of the five issues reported here, immigration draws the smallest share. It sits five points below democracy and twenty-two points below the economy. It still comes within a few points of two-thirds.
What Is Driving the Gloom
Ask Latino voters directly about the president’s effect on the economy and the answer comes back flat and specific. Seventy-four percent say President Donald Trump’s policies have made economic conditions worse. [1] That share was 61 percent in 2025. The figure is an average for the whole group, not a breakdown by party. Thirteen points in a year is a large move on a question that names a policy author. The question does not ask whether the economy is growing or shrinking; it asks who is responsible for the direction. The share naming the president as a cause of worse conditions grew by double digits, larger than the seven-point slide in the top-line rating of the economy.
The most visible cost in the survey is energy. The price of gas and energy is now one of the top economic concerns among Latino voters ahead of the midterms. Eighty percent say they are very concerned about it. In January, that share stood at 48 percent. Thirty-two points in roughly nine months. That is the largest single move among the trend lines reported here. The distance between 48 percent and 80 percent is wider than the distance on any other cost question in this survey. Fuel prices reach a household through the car and through the cost of heat and electricity. They also reach it weekly rather than once a year. The survey asks about concern, not about what any one family spends. So the figure measures how loud the subject is, not how large a single bill is. The question offers a scale, and these figures count only the respondents standing at the top of it.
Unemployment has moved in the same direction, more slowly and from a lower starting point. Fifty-nine percent of Latino voters now say they are very concerned about unemployment. The figure was 47 percent in September 2025. In January 2024, it was 40 percent. Three readings across roughly thirty-three months, each higher than the last, with no reversal in between. Nineteen points separate the first reading from the third. Twelve of those points arrived in the last year. A labor market that looks steady in aggregate can still register as a risk to a household. A worker who watches hiring slow at the plant or the warehouse where a relative works does not need a national unemployment rate to spike. Concern about a job requires only the possibility of losing one. The comparison here is between shares of a similar group at three points in time, not between three different questions. Three readings do not prove a trend on their own, but they do rule out a single bad month. Steady, one-directional movement is harder to dismiss than a spike.
Two kinds of question sit behind these numbers, and they should not be read as one. The issue ranking asks how important a subject is to a vote. The cost questions ask how concerned a person is. The first uses “very important” as its highest option. The second uses “very concerned.” Importance and concern are separate measures. So 86 percent on the economy and 80 percent on gas and energy neither cancel nor confirm each other. A voter can rank an issue first and still rate a specific cost second. Anyone comparing those two figures should hold the distinction in mind, because reading them as one measurement is the most common way to misread this survey. What the two families of questions do share is the direction of travel. On the issue that both coalitions rank first, the two coalitions still read present conditions differently. The survey records one agenda and two verdicts. That is the shape of this electorate in the fall of 2026. The same split appears more sharply on immigration enforcement.
Enforcement Rose and the Money Followed
Immigration enforcement is the subject on which the survey’s two halves separate most cleanly. Immigration-related arrests rose over the summer. They reached a new high for the current Trump administration, which is a comparison within a term rather than across decades. That rise followed large increases in immigration enforcement funding approved by Congress. The money was approved in two steps, one earlier in the summer and one in 2025. The sequence is worth keeping straight. The appropriation came first, and the arrest counts climbed after it. A budget line is the capacity to act. Agencies hire, train and deploy against the figure Congress writes. The count of arrests is the record of that capacity being used. Arrests measure enforcement activity rather than the size of any population. Funding decisions are made in one place, by one set of actors, on their own schedule. The consequences are counted somewhere else, by different officials, later. That lag is why a policy debate and its visible results rarely arrive in the same news cycle.
Against that backdrop, most Latino voters are critical of the administration’s approach to deportations. Sixty-two percent say the administration is doing too much when it comes to deporting immigrants who are living in the U.S. illegally. [1] Twenty-two percent say it is doing about the right amount. Fourteen percent say it is doing too little. Those three answers account for 98 percent of the group, and the rest chose none of the options offered. The direction of the criticism is clear from the size of the first number. The dominant complaint is not that enforcement is too slow. It is that enforcement has gone too far. Sixty-two percent is above the halfway line, which makes it a majority rather than the largest of several minorities. The survey records how large that majority is, not how strongly any one respondent holds the view.
The average conceals a clean split inside the Latino electorate. Latino voters who backed their Republican congressional candidate largely support the administration’s approach to deportations. Latino Democratic voters go the other way. Eighty-six percent of them say the administration is doing too much. The two readings sit at opposite ends of the question rather than on either side of a middle. On the economy, the two coalitions were closer. They were 11 points apart on how much the issue mattered to their vote. They divided 89 percent to 78 percent on which candidate they back. On deportation, the distance between them is wider than any gap the issue list produced. The report also finds that Latino registered voters view the Democratic Party more favorably than the Republican Party. That party-level reading sits alongside the issue-level split rather than replacing it.
The economy, at 86 percent, is what Latino voters say their vote is about, and it leads the list for the supporters of both parties. The verdict underneath it is darker: 19 percent called national economic conditions excellent or good, 74 percent said the president’s policies had made them worse, 80 percent said they were very concerned about the price of gas and energy, and 59 percent said the same about unemployment. Those readings were taken between Sept. 28 and Oct. 4, 2026, in the weeks before the November midterms rather than after them. A survey of this kind is not a forecast. It is a timestamp.
