Polls Show Democrats Leading National Ballot

A person using an electronic voting machine at a polling station — Photo: Lisa F. Young / Shutterstock

Conservative host Larry O’Connor says some 2026 midterm polls are being used to shape voter behavior, not measure it — as several national surveys show clear Democratic leads.

Story Snapshot

Larry O’Connor claims “weaponized” polls aim to depress Republican turnout.
Major national polls show Democrats leading the generic ballot by 8–12 points.
A recent fake-poll scandal shows how easy it is to publish bogus numbers, though experts say outright fakes are rare.
Shifts from registered-voter to likely-voter models may widen Democrats’ polling edge this fall.

What O’Connor Alleged And Why It Resonates

Conservative talk host Larry O’Connor argued that some high-profile polls are being “weaponized” to create a story of a blue wave and to drain energy and money from Republican campaigns. He urged conservatives to ignore headline polling and keep organizing, saying “the math isn’t mathing” in some races. Townhall summarized his remarks and described his warning as a push to fight, not sulk, in the final stretch before Election Day.

O’Connor’s claim taps into a wider fear that media and political elites push narratives that favor their side. Many Americans across the spectrum think insiders game the system. That feeling grew after real cases where fake polls spread online and reached news feeds. Those events make people doubt polling headlines and worry about turnout effects, even if the hard proof of coordinated suppression is thin.

What The Polls Actually Show Right Now

National surveys in recent weeks show Democrats ahead in the generic congressional ballot. The Marist Poll reported a 12-point Democratic lead among registered voters in mid-September. The poll also found that views of President Trump loom over the midterms, shaping voter choice down ballot.

Quinnipiac University’s national release found Democrats up 11 points on control of Congress, reinforcing the narrative of a sizable lead as the calendar moves closer to Election Day.

Why Different Polls Can Disagree

Polling methods vary. Some pollsters interview registered voters, while others switch to likely-voter models as voting nears. That shift can change the numbers. Yahoo’s roundup noted that after Labor Day, more firms moved to likely-voter screens, and that change helped increase the measured Democratic edge in some series by a couple of points.

Response rates, weighting choices, and how pollsters model the electorate also matter. Even seasoned pollsters say low response and complex turnout patterns make exact reads tough. That is why single polls can mislead, and averages or trends tend to be more useful than outliers or viral charts.

The Fake-Poll Problem — Real, But Not The Norm

In August, a group calling itself Median Strategies admitted it fabricated polls as a “short-term social experiment.” The Los Angeles Times and The New York Times reported that the phony numbers still reached real audiences before being retracted. Experts said the episode shows how easy it is to publish bogus polls online, which can then shape coverage or social chatter.

Five weeks out from the midterms, here’s the state of things as I see it.

Polling averages clearly indicate that the Democrats will pick up at least one statewide seat here in Texas. If it’s a tsunami, it might be more.

These races, particularly the Paxton/Talarico race, have…

— John Huffman (@JohnRHuffman) September 29, 2026

PBS NewsHour highlighted the same risk and warned that vetting poll sources is key. While outright fake polls are rare, they can still bend a news cycle and public mood. The lesson is simple: check the sponsor, sample, method, and track record before sharing or believing a big, sudden swing.

How Voters Should Read The Moment

Voters face two truths at once. First, most credible national polls point to a Democratic lead of mid-to-high single digits or more. Those are public snapshots, not guarantees, but they are consistent across several respected firms. Second, claims of “psyop” polls should be handled with care. Bad actors do exist, but strong evidence of a broad, coordinated plot remains scarce. The safest move is to vote, not to let headlines decide whether you do.