America’s Midterms: The political ground is shifting, but how far?

Editor’s notes:

1. This article is deliberately cautious about forecasting. It distinguishes polling results, race ratings and campaign behaviour rather than treating them as interchangeable evidence.
2. All figures and dates are based on reporting available as of October 19, 2026.
3. For context and comparison, readers may also wish to consult our earlier article on the midterms, published October 15, 2026.

With the November 3 midterms approaching, new polling, changing House race ratings and the battle for the Senate offer fresh clues about the American political mood. But a shift in sentiment is not the same thing as a change in power.

There is a difference between a political party having a bad few weeks and a country deciding it has had enough.

The American midterms are beginning to offer clues about which of those things might be happening.

When we last examined the early signs of a political shift, the important question was whether the numbers emerging from the campaign trail would translate into something more substantial. With the election now approaching, we have a little more evidence to work with.

The latest developments are significant. But they also come with qualifications, and it is worth keeping the two in perspective.

The House: More Republican seats are in play

The clearest recent development has been the changing assessment of competitive House races.

On October 16, the Cook Political Report moved 11 House races towards the Democrats. Among the changes, Montana’s First Congressional District, which Donald Trump carried by 12 percentage points in 2024, was downgraded from leaning Republican to a toss-up. Several other Republican-held seats were also reassessed as more competitive.

These are not election results. Cook’s ratings are assessments of how competitive individual contests have become, taking account of factors including polling, candidate strength and the political environment.

Nevertheless, the changes matter. They suggest that some seats previously regarded as relatively comfortable for Republicans are now attracting greater scrutiny.

The House has 435 voting members, and 218 seats are normally required for a majority when all seats are filled. A party does not need to win the national popular vote by a landslide to change control of the chamber. It needs to win enough of the contests that matter.

That is precisely why individual districts deserve attention.

A national poll can tell us something about the mood of the electorate. It cannot tell us, on its own, how that mood will be distributed across hundreds of separate contests.

Source:

• Washington Post, Cook Political Report moves 11 more House races toward Democrats

The Senate: A different electoral calculation

The Senate presents a rather different challenge.

Republicans currently hold a 53–47 majority, counting the independents who caucus with the Democrats on the Democratic side. Only around a third of Senate seats are contested in a typical election cycle, meaning that the party seeking control must navigate a much narrower set of opportunities.

Democrats have competitive races to fight, but some of their targets are in states that Trump won comfortably in 2024. That creates an obvious problem: a national shift towards the Democrats does not automatically translate into victories in states with a strong Republican voting history.

At the same time, several contests are attracting considerable attention. Georgia, Maine, North Carolina, Ohio and Texas are among the states featuring prominently in the current Senate battle, although the circumstances differ from one race to another.

The financial stakes are substantial, too. Republican spending advantages and the concentration of campaign resources in battleground states mean that polling leads should not be treated as guarantees.

The Senate is therefore a useful reminder that electoral geography matters. A party can improve its standing nationally and still find that the seats it needs are difficult to win.

Source:

• Reuters, The races that will decide control of the US Senate

Trump remains central to the argument

It is difficult to separate the midterms from the presidency.

A Washington Post–ABC News–Ipsos poll published on October 18 put Trump’s approval rating at 33 per cent, with 66 per cent disapproving. The same poll found Democrats leading Republicans by eight percentage points on the question of which party respondents preferred to control Congress.

Those are substantial figures, but they need to be read carefully. Approval ratings measure views of the president; a congressional preference question measures something different. Neither tells us precisely how people will vote in each individual race.

Other polling has also pointed to economic concerns, including inflation and the cost of living, as important issues for voters. The economic consequences of the war with Iran have added another dimension to the debate.

For Republicans, the difficulty is that their candidates must campaign on their own records while defending a party closely associated with the president. Some candidates may benefit from Trump’s support among his core constituency. Others may find that association makes it harder to appeal to voters who are dissatisfied with the administration.

That is a political dilemma, not a verdict on the election.

The question is whether dissatisfaction with the president will be strong enough, widespread enough and concentrated in the right places to affect the balance of power in Congress.

Source:

• Washington Post, Poll finds jump in Republicans who say Trump is enriching himself

Early voting: What can we actually conclude?

Early voting remains another useful piece of evidence, but it is one that requires restraint.

An increase in early ballots does not automatically favour one party. Nor does party registration necessarily tell us how an individual will vote. Early-voting rules vary between states, and comparisons with previous elections must account for differences in the states and populations being measured.

There is also the question of who votes on Election Day.

A campaign can appear energetic online, attract large crowds and dominate the news cycle without necessarily bringing enough supporters to the ballot box. Conversely, a quieter campaign may have a more reliable turnout operation.

The distinction between enthusiasm and participation is particularly important in a midterm election, when turnout is generally lower than in a presidential election.

Early-voting figures should therefore be treated as a snapshot of participation, not a preview of the final result.

The same caution applies to polls. They are measurements taken at a particular time, with margins of error and methodological limitations. They can identify changes in public opinion, but they cannot remove the uncertainty inherent in an election that has not yet happened.

A question of trust

There is another development worth watching beyond the familiar contest over seats and polling.

A Reuters survey published on October 15 examined whether candidates in competitive Senate races would commit to accepting the election results regardless of the outcome.

Of the 26 campaigns approached across 13 states, only 14 gave unconditional assurances. Nine of those were Democrats, one was an independent and four were Republicans. Other campaigns offered conditional responses or did not answer.

That finding does not establish that an election will be disputed, nor does it establish that any particular candidate intends to challenge a legitimate result. It does, however, demonstrate that acceptance of election outcomes remains an issue in American politics.

For a country whose political institutions depend on the peaceful transfer of power, that is not an insignificant question.

The counting of votes, the certification of results and the willingness of candidates to accept legitimate outcomes are fundamental parts of an electoral system. They deserve scrutiny independently of which party wins.

Source:

• Reuters, Just over half of candidates in key US Senate races say they will accept results, others hedge

What happens next?

The remaining days of the campaign will provide further opportunities to test the current picture.

Will the movement in House race ratings continue? Will polling in the Senate battlegrounds become more favourable to either party? Will campaign spending reveal where the parties believe their vulnerabilities lie? And will the voters who tell pollsters they are dissatisfied actually turn out?

These are more useful questions than trying to declare a winner before the votes have been counted.

There are also reasons to be cautious about predicting a sweeping result. Partisan divisions remain deep, the number of genuinely competitive House races is limited, and the Senate’s electoral map creates obstacles that national polling cannot fully capture. Past polling errors are another reason not to treat today’s numbers as tomorrow’s results.

The November 3 election will provide the answer. Until then, the evidence is best understood as a changing set of indicators rather than a settled outcome.

The ground is moving. The destination remains uncertain.

There is enough new information to suggest that the political environment has become more challenging for Republicans in a number of important contests. The latest House ratings, the polling on congressional preference and the president’s approval figures all contribute to that assessment.

But there is a difference between identifying a political trend and assuming it must produce a particular result.

The House and Senate operate under different electoral conditions. Individual candidates have their own strengths and weaknesses. Turnout can alter the picture, and the distribution of votes matters as much as the national mood.

For now, the most defensible conclusion is also the simplest: the political ground is shifting, but how far it will shift remains to be seen.

On November 3, Americans will have their say. The rest of us can watch the evidence emerge, distinguish the signals from the noise and wait for the votes to tell their own story.