Before the Sweep: 20 Months of Trump Polling

Change Research's Compass Poll national surveys, conducted from February 2025 through October 2026, show a 10-point drop in Trump's job approval alongside a widening Democratic lead on the generic congressional ballot. In the most recent wave, conducted October 12–18, 2026, among 2,240 registered voters, 37% approved of Trump's job performance and 62% disapproved, a net of -25. In the earliest wave, conducted in February 2025, 47% approved and 52% disapproved, a net of -5.

The 2026 midterm elections, held November 3, were the first major electoral test of Trump's second term. Democrats needed a net gain of three seats to recapture the House majority from a narrow Republican majority. According to NBC News projections, Democrats won control of both chambers, flipping seven Senate seats and winning at least 258 House seats, well beyond the 218 needed for a majority. Early exit polls put Trump's approval at 38%, closely matching the 37% recorded in the final Compass Poll wave, and 89% of voters said the cost of living was a major or crucial factor in their vote. Twenty months of Compass Poll tracking show how voter opinion moved in the run-up to that result.

Trump's Approval Fell 10 Points

Among registered voters, Trump's approval fell from 47% in February 2025 to 37% in September and October 2026. The steepest drop came between May and June 2026, when approval fell from 43% to 39% and disapproval rose from 56% to 60%. The decline came even as Republicans stayed loyal: between 87% and 94% approved of Trump's job performance in every wave.

Independents were far more negative. In September 2026, just 14% of self-identified independents approved of Trump's job performance and 81% disapproved. Their approval ticked up to 17% in October, with 78% disapproving.

Trump's approval on border control and immigration – arguably his signature issue heading into his second term – held up better than his overall rating. Net approval briefly turned positive in April 2026, at +1, before turning negative again in September, the next wave to include the question. In October 2026, 47% approved and 51% disapproved of his handling of border control, a net of -4.

Democrats' Generic Ballot Lead Grew to 11 Points

In the February 2025 wave, 47% of voters selected a Democratic congressional candidate and 45% selected a Republican, a Democratic advantage of 2 points. Democrats' strongest 2025 readings came in August, October, and November, when the margin reached about 4 points. However, the advantage narrowed to less than 1 point in December 2025, when 43% of voters selected a Democratic candidate and 43% selected a Republican.

The Democratic margin widened in 2026, holding between 3 and 8 points from March through September before reaching 11 points in October. By October 2026, 50% of voters selected a Democratic candidate and 39% selected a Republican, a Democratic advantage of 11 points. Both the Democratic share and the margin were the highest in the tracking window.

Among self-identified independents, 43% selected a Democratic congressional candidate in October 2026, compared with 16% who selected a Republican, a 28-point Democratic advantage among that group.

Cost of Living Was Trump's Weakest Issue

The cost of living was Trump's lowest-rated issue in every wave that asked about it. In October 2025, the first wave in the series to include the question, 37% approved and 60% disapproved, a net of -22. Disapproval rose through 2026, reaching 70% in September, when 28% approved, a net of -42. In October 2026, approval edged up to 30%, with 67% disapproving, a net of -37.

Among Republicans, cost-of-living approval fell from 77% in February 2025 to 62% in September 2026, a decline of 15 points within his own party.

By the final wave before Election Day, Trump's approval stood at 37%, Democrats led the generic ballot by 11 points, and 67% of voters disapproved of his handling of the cost of living. Exit polls on November 3 pointed in the same direction.

Methodology: The most recent wave in this tracking series was conducted online from October 12–18, 2026. Using Dynamic Online Sampling to attain a representative sample, Change Research polled 2,240 registered voters nationally. The modeled margin of error for that wave is ±2.3 percentage points. This article draws on a series of Compass Poll national surveys conducted from February 2025 through October 2026; individual wave sample sizes ranged from approximately 1,300 to 2,700 registered voters nationally, with modeled margins of error ranging from ±2.0 to ±2.9 percentage points. As a member of AAPOR's Transparency Initiative, Change Research discloses all relevant details about our research, with the principle that the public should be able to evaluate and understand research-based findings in order to instill and restore public confidence in survey results.