Are the midterms important for Trump? And for the US?

In the midterm elections, all 435 members of the House of Representatives and 35 of the 100 senators will be up for election (see the first two charts).1 The Democrats need to increase their number of representatives from 214 to 218 to regain the House, while they require four additional seats to achieve a majority of 51 senators, as the vice presidency – and therefore the tie-breaking vote – will remain in Republican hands.

The extent of a Democratic victory will depend on which chamber is won. Control of the Senate would allow them to block appointments of federal judges, senior administration officials, or members of the Federal Reserve. Control of the House would grant them greater powers to conduct investigations into the Executive, demand documents and examine potential conflicts of interest or abuses of power – a strategy which Democratic leaders seem to prefer over immediately initiating a new impeachment process. A Democratic majority in either chamber could block new Republican legislation, while control of both would allow an alternative agenda to be advanced, although Trump would retain veto power. In the fiscal sphere, the new political balance could modify some of the planned spending cuts in the One Big Beautiful Bill Act (OBBBA) or make it difficult to extend its temporary provisions, although it would be unlikely to alter substantially the projected path of the national deficit and debt in the absence of bipartisan agreements. The political confrontation could also extend to funding bills, complicating the approval of federal spending and increasing the risk of partial government shutdowns, with the associated economic costs.2

Losing control of Congress would not prevent Trump from continuing to act through executive orders and powers delegated by previous laws, especially on matters related to migration, trade and foreign policy. He would have considerable room for manoeuvre, though not unlimited: he would need funding to implement some measures and his decisions would remain subject to existing laws and court oversight. The Supreme Court’s setbacks to the tariffs imposed under the IEEPA and the attempt to restrict birthright citizenship illustrate these limits.3

The Democrats have history on their side: the president’s party has lost seats in the House in 20 of the 22 midterm elections held since 1938.4 This consistency does not determine the outcome, but it reinforces the interpretation of the opinion polls, which already place the Republicans in a vulnerable position.

At the end of September, the average of the polls on congressional voting intentions placed the Democrats some 7-8 points ahead of the Republicans (see third chart). This advantage does not automatically translate into seats, given the district boundaries and the geographical distribution of votes, but it is consistent with a significant risk to the narrow Republican majority in the House. The prediction market platform Polymarket assigned, at the end of September, a 90% probability to a Democratic House and just under a 40% probability to the Republicans retaining the Senate (see fourth chart). The difference is due to the characteristics of the electoral map. To regain control of the Senate, the Democrats need to emerge with a net gain of four seats. Their main opportunities lie in states currently represented by Republicans, such as North Carolina, Maine, Alaska and Ohio, while Iowa and Texas present more challenging options. At the same time, they must retain Georgia and Michigan, two states won by Trump in the 2024 presidential elections.5

Elections results have offered similar signals. In 2025, the Democrats took control of the government in Virginia, retained power in New Jersey, and maintained the progressive majority in the Wisconsin Supreme Court (whose members are elected at the polls) in a state where Trump had won in 2024. More revealing still is the pattern of elections held since his inauguration: in an analysis published in April on 229 federal and state elections with candidates from both parties (mainly special elections for Congress and state legislatures), Democrats exceeded the vote share obtained by Kamala Harris in 2024 in 193 of them, with an average improvement of 5 points.6 In the New York mayoral elections, Zohran Mamdani’s victory, with a campaign focused on housing and the cost of living, provided further evidence that an alternative economic agenda could gain support even in neighbourhoods where the Republican vote had made significant gains in the last presidential elections.

Trump’s victory in 2024 was supported by greater mobilisation among Republican voters, the abstention of former Democratic voters, and concerns about inflation and immigration. His gains were widespread across the country: his vote share increased compared to 2020 in 9 out of 10 counties and gained ground most notably in large urban areas, regions with a high immigrant population, southern Florida, the Texas border region, and the industrial states of the Midwest.7 The question now is whether those areas have experienced the changes they were hoping for and whether their voters feel their situation has improved.

The available data do not allow us to identify any distinct economic dividend in the areas that led the swing towards the Republicans. The US economy has maintained positive growth, but in August 2026 only eight states recorded a statistically significant year-on-year increase in non-farm employment, while 10 showed notable rises in their unemployment rates compared to a year earlier.8 These statistics indicate that the improvement in employment is far from widespread, although they do not allow a causal link with Trump’s policies to be established. It is also worth noting that voters’ perceptions are influenced by party identification.9

Tariffs provide protection against foreign competition for producers of steel and other goods, but they also increase costs for companies that use imported metals and components. This tension is particularly relevant for the industrial belt of Michigan, Pennsylvania, Wisconsin and Ohio. The available evidence suggests that tariffs have increased the cost of imported goods without yet producing a noticeable improvement in overall employment, with modest and uneven labour effects across sectors and regions.10 Moreover, the energy shock triggered by the conflict in the Middle East has added new inflationary pressures, directly affecting fuel costs, which are one of the most visible prices for American households. It has also driven up interest rates, complicating the main economic expectation placed on Trump: bringing down the cost of living.11

Migration policy presents a similar tension. Some preliminary analyses link the intensification of operations and detentions carried out by the Immigration and Customs Enforcement (ICE) agency with declines in local labour supply, consumption and employment, including that of US-born workers.12 Their impact could also be electoral. Trump significantly increased his support among Latino voters in 2024, largely driven by economic discontent, but the 2025 elections in Virginia and New Jersey reversed some of that progress. Polls indicate that his support is weakening amid high prices and these communities’ exposure to ICE operations, although it is still unclear how much is due to changes in voting and how much to increased mobilisation among Democrats.13