Luntz Says Polls May Be Missing Trump Voters
Frank Luntz is warning that current polling may be asking the wrong people the wrong questions. Appearing on CNN News Central, Luntz was asked whether surveys are under-sampling the voters Donald Trump tends to bring out. He said Trump voters often ignore calls because they believe the political system is rigged against them. That creates a familiar problem: a poll can look precise while missing voters who are less likely to answer it. Luntz said he got 2016 wrong after trusting exit polls that projected Hillary Clinton would defeat Trump in a landslide. He said many Trump voters walked away from polling because they believed the surveys were biased. That is not exactly a tiny footnote when turnout can decide an election.
Turnout Matters More Than a Simple Survey
Luntz said turnout is the statistic that matters most, yet he is not convinced polling firms measure it properly. He argued that many surveys ask how the electorate feels instead of identifying people who will actually participate. A voter may say they plan to vote, but that answer alone does not tell pollsters whether the person is paying attention or will show up. Luntz said focus groups and one-to-one interviews can help separate engaged voters from people who are likely to stay home. His point is simple: counting every stated preference the same way can produce a polished picture of an unreliable electorate. Elections are decided by people who cast ballots, not merely by people who sound interested on a phone call.
Three Questions Pollsters May Be Skipping
To judge turnout, Luntz urged viewers to watch how polls ask their questions. He named three items he considers essential: how interested a person is in the election, how often that person follows election or political news, and how often the person talks politics with friends and neighbors. He said many polls do not ask those questions. He also warned that people can lie when asked whether they will vote, so a simple yes is not enough. Luntz said deeper questions can reveal who is engaged and who is merely being counted. His skepticism is personal as well as professional. When asked how he could acknowledge frustration among Trump supporters while still doubting the midterm forecast, he said he had to remain skeptical because he had once been so convinced that he was right, only to get it wrong.
Even CNN Sees Room for a Republican Turnaround
CNN chief data analyst Harry Enten offered a similar warning earlier this month. He said Republicans have a substantial chance of keeping their Senate majority if polls are off by only a few points. Enten pointed to a possible 3.5-point undercount of Republicans, saying that size of error is possible based on what happened in 2020 and in several key states in 2024, including Texas. He also cited the early October generic congressional ballot, which he said had underestimated Republicans in 14 of 16 elections since 1994, by an average of four points. A CNN scenario showed Republicans at 52 Senate seats and Democrats at 48, with Michigan marked as a Republican flip and Maine, Texas, and Ohio marked as Republican holds. Those figures are a scenario, not a result, but they show why turnout and polling methods matter.
WE’D LOVE TO HEAR YOUR THOUGHTS! PLEASE COMMENT BELOW.
CNN: If early October polls keep the pattern of underestimating Republicans, they will keep the senate
🔴 Republicans: 52
🔵 Democrats: 48States:
Michigan: 🔴 Flip
Maine: 🔴 Hold
Texas: 🔴 Hold
Ohio: 🔴 Hold https://t.co/11RAUtcLAe pic.twitter.com/dVUU90nSHj— OSZ (@OpenSourceZone) October 12, 2026