National midterm polling has become an early measure of whether either party can build a durable advantage before the 2026 congressional elections. The latest Harvard CAPS/Harris polling series shows that edge changing since spring, with Democrats holding a small lead in March before the race moved into a tie by April and stayed competitive through the summer, according to releases from the poll. The results point to an unsettled electorate rather than a decisive break for either party.
Democrats’ early edge narrowed as the race moved to even
The Harvard CAPS/Harris Poll reported on March 30 that Democrats held a 51% to 49% advantage on the congressional ballot among voters overall, with the lead widening to four points among likely midterm voters, according to the poll’s March press release and topline. The March survey was conducted March 25 and 26 among 2,009 registered voters by The Harris Poll and HarrisX. In the same release, poll officials said 73% of respondents said they would definitely or probably vote in the 2026 midterm elections.
By April 28, that margin had disappeared in the same polling series. The April Harvard CAPS/Harris Poll found the congressional horserace split 50-50 among both general voters and likely midterm voters, according to the poll’s April press release. That survey was conducted April 23 through 26 among 2,745 registered voters.
The poll archive on the Harvard CAPS/Harris site indicates that the race remained tied in July, with the July release summarizing the congressional horserace at 50-50. The archive also shows an August 2026 poll was posted with key results, though a press release was not listed in the archive snapshot reviewed. Taken together, the series shows a shift from a narrow Democratic advantage in late March to a dead-even national ballot by late spring and summer.
The poll is national, not state-based, so it offers a broad measure of partisan mood rather than a district-by-district map of House or Senate control. Harvard CAPS/Harris has not released, in the materials reviewed here, a comprehensive state breakdown showing where the Democratic March edge weakened the most. That means the polling does not by itself identify which battleground states or congressional districts are driving the change.
What is confirmed is that independent voters were a key part of the picture in April. The April release said 52% of independents would vote for a Democrat if the election were held at that time, even as the overall ballot stood at 50-50. That split suggests each party was retaining different sources of strength within the electorate rather than one side consolidating a broad national advantage.
The turnout measures in the poll also remained closely watched. In March, 73% said they would definitely or probably vote in the midterms, while in April the share saying they planned to vote was 71%, according to the two releases. Republicans and Democrats each posted 62% in “definitely vote” responses in March, while independents were substantially lower, underscoring how turnout composition could shape the eventual result.
The clearest through-line in the polling is the economy. In the March release, 45% of voters said their personal financial situation was getting worse, and 71% said inflation was above 3%, while 35% said the country was on the right track. Mark Penn, co-director of the Harvard CAPS/Harris Poll and chairman and CEO of Stagwell, said at the time that gas-price increases were likely behind President Donald Trump’s approval decline and the tighter political environment.
The April poll reinforced that message with more detail on costs. It found 57% of voters believed the economy was shrinking, 73% said inflation was above 3%, 88% noticed higher gasoline prices and 84% noticed higher grocery prices. It also found that 85% were concerned higher gas prices would lead to broader increases in the cost of living.
For voters, the practical takeaway is that the midterm landscape remained highly fluid through at least the summer polling in this series. The Harvard CAPS/Harris data does not show either party with a locked national advantage after March; instead, it shows a competitive race shaped heavily by economic sentiment, turnout intent and perceptions of day-to-day costs. As of the latest releases available from the poll archive reviewed, the congressional ballot remained closely divided.