The national picture days before the vote
Days before voters go to the polls in the 2026 midterm elections, President Donald Trump is in positive net approval territory in only eight of the 50 states, according to the latest state-by-state tracker from Civiqs reported by USA Today on October 31, 2026. In the other states, more voters disapprove of the president’s job performance than approve of it, or the two sides are tied, leaving the White House facing a difficult political map at the exact moment when control of Congress is about to be decided.
The finding does not, by itself, decide any election. A presidential approval rating is a measure of how voters feel about the president, not a ballot for a House or Senate candidate, and state-level results are shaped by local candidates, local issues, turnout, and the particular choices on each ballot. But approval ratings are closely watched because they often describe the mood in which those local races are fought, and because they can affect enthusiasm, fundraising, and the willingness of voters in each party to turn out.
The latest national picture points in the same direction as the state map. The Civiqs tracker, based on a large rolling survey of registered voters collected since the start of the president’s second term, put national approval at 34 percent and disapproval at 60 percent in its October 30, 2026 update, with 6 percent neither approving nor disapproving. A separate national survey from Emerson College Polling, conducted October 24 to 26 among 1,000 likely voters, put approval at 37 percent and disapproval at 57 percent. Different pollsters use different methods, samples, and questions, so their exact numbers differ, but the broad direction across the major averages in late October was consistent: the president was well underwater nationally, with disapproval close to or above 60 percent in several averages.
That context matters because midterm elections are often read, fairly or not, as a verdict on the sitting president. The party in the White House has historically tended to lose ground in midterms, and a president with approval in the high 30s faces a steeper climb when asking voters to keep his party in control of the House and the Senate. Supporters of the president, by contrast, have pointed in past cycles to polling misses and have argued that surveys can understate his support. Polling analysts themselves caution that a poll is a snapshot of opinion at the time it was taken, subject to sampling error and to shifts in the final days of a campaign, and that it should not be treated as a prediction of the final result.
Trump approval rating by state: the eight states still in positive territory
The eight states where the Civiqs tracker shows Trump with a positive net approval rating are Alabama, Arkansas, Idaho, North Dakota, Oklahoma, Utah, West Virginia, and Wyoming. USA Today reported that all eight are states the president carried in his past three presidential campaigns, which underlines how narrow his remaining geographic base of net-positive approval has become in this tracker.
Wyoming remains the president’s strongest state in the tracker, with a net approval rating of plus 14 percent. That is a notably smaller margin than the president enjoyed in Wyoming at the start of his second term, when earlier Civiqs updates put his net standing far higher, and it illustrates a broader pattern in the data: approval has fallen in every state since January 2025, even in the states where it remains positive. In other words, the eight-state list is not a sign of stable strength in those states so much as a sign that the decline there has not yet pushed the balance below zero.
Kansas, meanwhile, sits exactly even in the latest USA Today account, with 46 percent approving and 46 percent disapproving of the president’s job performance. An even split is not a positive rating, and it is a striking result for a state that has long been reliably Republican in presidential elections. The same tracker has at various points in October placed other traditionally Republican states, including Montana and South Dakota, at or near even, a reminder that the movement in the data is not confined to Democratic-leaning states.
For readers trying to interpret a net approval figure, the arithmetic is simple: net approval is the share who approve minus the share who disapprove. A state at plus 14 percent, like Wyoming in the latest update, means approval exceeds disapproval by 14 points, not that 14 percent of voters approve. A state at minus 25 percent means disapproval exceeds approval by 25 points. The size of the gap, rather than the approval figure alone, is what the tracker is designed to show, and it is why a state can have approval in the 40s and still be deeply negative on net if disapproval is higher still.
The swing Senate states where Trump is underwater
The most consequential part of the state map for the battle for the Senate is not the eight states where the president remains positive, but the states where competitive Senate races are being fought and where he is negative. According to the Civiqs tracker reported by USA Today, Trump has a negative net approval rating in all six states described as swing Senate states: Alaska at minus 25 percent, Michigan at minus 26 percent, Maine at minus 29 percent, North Carolina at minus 18 percent, Ohio at minus 18 percent, and Texas at minus 20 percent.
Those figures do not mean that every Republican Senate candidate in those states is trailing, and they do not mean that every Democratic candidate is leading. Senate races are decided by the candidates on the ballot, by their campaigns, and by voters’ views of those candidates, not by the president alone. But a president who is 18 to 29 points underwater in a state is not, in polling terms, an asset that a candidate in that state can easily lean on, and campaigns in both parties will be reading the same numbers as they decide where the president should campaign, where he should stay away, and which messages to emphasise in the final days.
The pattern extends beyond the six swing states. USA Today also reported negative net approval in reliably Republican states such as Kentucky, at minus 13 percent, and South Carolina, at minus 10 percent. When a president is underwater in states his party normally wins comfortably, it can signal weaker enthusiasm among his own base, a greater willingness among some normally loyal voters to stay home, or a readiness to split a ticket. It can also reflect dissatisfaction with specific issues, such as the cost of living, which Emerson found was the top voting issue for 34 percent of voters in its final national survey, ahead of fighting corruption in government at 28 percent.
Taken together, the state numbers help explain why the closing days of the campaign have featured such intense attention on a small number of Senate contests. With the Senate majority potentially turning on a handful of seats, even a modest shift in turnout or in the choices of independent voters in states like Michigan, North Carolina, Ohio, or Texas could matter more than the national mood alone. The polls describe that mood; they do not replace the votes that will be cast on November 3.
Emerson College poll puts Trump approval at 37 percent, a new low
Separately, the final Emerson College Polling national survey before the midterms, released on Friday, October 30, 2026, found the president’s approval rating at 37 percent, the lowest he has recorded in Emerson College polling. That is a two-point drop from his previous low of 39 percent, a figure he had recorded in Emerson’s September, July, June, and May surveys. In the October survey, 57 percent of likely voters disapproved of the president’s job performance, giving him a net disapproval of 20 points.
The survey was conducted from October 24 to 26, 2026, among 1,000 likely voters, and Emerson reports a credibility interval similar to a margin of error of plus or minus 3 percentage points. That interval is important: it means the true level of approval in the population surveyed could plausibly be a few points above or below 37 percent, and that a two-point change from one survey to the next can sit within the range of sampling variation. What makes the October result notable is not a single two-point move, but that it is the lowest point in this pollster’s series for this president and that it continues a longer slide from the roughly even split, 46 percent approval to 47 percent disapproval, that Emerson recorded when its tracking began 15 months earlier.
Emerson’s executive director, Spencer Kimball, said in the pollster’s release that Democratic support had climbed eight points over that 15-month period, from 44 percent to 52 percent on the generic ballot, while Republican support ended the period at 42 percent, exactly where it had started. Kimball also noted that about 5 percent of voters remained undecided, most of them independents, and that about half of those undecided voters disapproved of the president, which he said could mean Democrats’ ceiling in the survey was higher than the headline figure suggested. That is an interpretation offered by the pollster, not a result, and it should be read as such.
Other polling averages reported in late October told a broadly similar story, even though their exact figures differed. As of October 30, the RealClearPolitics average put Trump at 38.4 percent approval and 59.7 percent disapproval, the New York Times average put him at 37 percent approval and 60 percent disapproval, and the Silver Bulletin average put him at 37.3 percent approval and 59.9 percent disapproval, according to USA Today’s summary. Averages smooth out the variation between individual polls, but they are still averages of polls, not votes, and they can move in the final days as new surveys are released.
Democrats lead by 10 points on the generic congressional ballot
On the question that comes closest to asking voters how they will vote for Congress, the Emerson survey found Democrats leading Republicans by 10 points, 52 percent to 42 percent, with 5 percent undecided. That is a wider Democratic margin than Emerson recorded in its October 2025 survey, when Democrats led by four points, and it continues a pattern in which the generic ballot has moved toward Democrats over the past 15 months even as Republican support has held roughly steady at the level it started from.
The generic ballot is not a forecast of House seats. It is a national question about party preference, and House control is decided in 435 separate districts, some of which are heavily partisan and some of which are competitive. A national lead can coexist with a close fight for the majority if the votes are distributed unevenly, and polling analysts note that the relationship between the national popular vote for the House and the seat count depends on district boundaries, candidate quality, and turnout in each district. Still, a 10-point lead is a substantial margin on a national question, and it is one reason that forecasters, analysts, and betting markets have been giving Democrats a better-than-even chance of winning the House.
The same survey also asked about a range of issues and about how voters viewed the president’s handling of them. Emerson reported that the economy was the top voting issue, chosen by 34 percent of respondents, followed by fighting corruption in government at 28 percent, immigration at 12 percent, health care at 11 percent, and threats to democracy at 10 percent. Those issue rankings help explain why both parties have spent the closing weeks of the campaign arguing about prices, jobs, and household budgets, even as they also argued about immigration, health care, and the direction of American democracy.
How to read these polls, and what they cannot tell us
Readers often ask how polls that give different numbers can all be right, and how a pollster can claim to measure the views of millions of voters by surveying about 1,000 people. The answer is that a well-conducted poll uses a sample designed to represent the larger population, weighting its results to match known characteristics of that population, such as age, sex, race, education, and geography. If the sample is genuinely representative, a sample of 1,000 can give a useful estimate, but it will always carry uncertainty, usually expressed as a margin of error or credibility interval, and it can be affected by who chooses to respond and who does not.
That is why analysts tend to look at averages and at trends across many polls rather than at a single survey. A single poll can be an outlier, a pollster can have a house effect that makes its results consistently lean one way, and the timing of a survey can matter if a major news event occurs during fieldwork. An average of recent polls smooths out some of that variation, but it does not remove the possibility that all polls share a common error, a risk that polling analysts have written about extensively after the misses of 2016, 2020, and, in a different way, 2024.
There is also a difference between a poll of registered voters and a poll of likely voters. Emerson’s October survey was of likely voters, meaning it tried to screen for people who are most likely to actually vote on November 3. Civiqs, by contrast, surveys registered voters on a rolling basis. Likely-voter models depend on assumptions about turnout, and those assumptions can be wrong, particularly in a midterm election when turnout is typically lower and more uneven than in a presidential election. That is another reason to treat polls as a guide to the state of the race, not as a substitute for the result.
Finally, approval ratings and ballot questions measure different things. A voter can disapprove of the president and still vote for a Republican House candidate, or approve of the president and still vote for a Democrat, for reasons that have to do with the local candidates, with the voter’s views on specific issues, or with a desire to check the power of whichever party holds the White House. The state approval numbers and the national ballot numbers in this story point in the same direction, but they are not the same measurement, and they should not be added together or treated as if one proves the other.
None of that makes polls useless. Polls remain one of the best available tools for understanding public opinion, and when many polls from different pollsters using different methods point the same way, that agreement is informative. What it does mean is that a responsible reading of the polls in the final days before a midterm is a cautious one: the president is unpopular by the measures available, Democrats lead on the national ballot question, and the state map is difficult for the White House, but the election will be decided by the voters who actually cast ballots on November 3, not by the voters who answered a survey in late October.
What happens next
Americans will vote on Tuesday, November 3, in elections for all 435 seats in the House of Representatives and for 35 seats in the Senate, along with a series of state and local contests. The results will determine whether Democrats take control of one or both chambers of Congress, whether Republicans hold on to the majority they currently enjoy in both, or whether control is split. Polling in the final days will continue to be released, and campaigns will continue to argue that their internal numbers tell a different story, but the only numbers that will matter after November 3 are the votes themselves.
For the White House, the stakes are unusually high. A Democratic House would be able to block much of the president’s legislative agenda, to launch investigations, and to use its subpoena power, while a Democratic Senate would be able to block confirmations and to shape the final two years of the president’s term. For Democrats, the stakes are equally clear: a failure to win back at least one chamber after leading in the polls for months would be seen as a major disappointment, and it would raise difficult questions about the party’s message and its candidates. For voters, the stakes are the more immediate ones that the polls themselves identified: the cost of living, the direction of the economy, and the future of the country’s democratic institutions.
Related USN coverage includes reporting on Democrats opening leads in four Senate battlegrounds and on Trump and Obama trading blows on the campaign trail.
Sources
USN reviewed reporting from multiple outlets for this story. Read more from USA Today, Emerson College Polling and Civiqs.