The polls say Democrats are headed for a wave in the midterms. But pollsters have been wrong before, and it wouldn’t take much of a miss to flip that script.
That’s the worry hanging over races nationwide, including marquee Senate contests in red states like Ohio and Texas. Democrats hold narrow leads in public surveys in those GOP strongholds and are counting on wins to take back the upper chamber.
It all comes down to a bigger question: Can the polls be trusted?
People were interested in these podcasts
Pollsters have taken plenty of heat after several high-profile misses in recent cycles, and many analysts say the job is only getting harder. Fewer people pick up calls from unknown numbers, response rates keep dropping, and campaigns are sifting through scattered streams of data to figure out who will actually show up on Election Day.
CNN data analyst Harry Enten said even a modest miss in Republicans’ favor could scramble expectations. It could flip Democratic-held Senate seats such as in Michigan and even help Republicans hold a House that’s historically been tough for a president’s party to keep in the midterms.
“If the polls underestimate Republicans by two to four points, well, remember all that blue that was on the map before? All that blue went bye bye, see you later, adiós, amigos, sayonara,” Mr. Enten said.
SEE ALSO: Trump on midterm stakes: ‘Just look what’s happening in France’
Matt Towery, the GOP pollster who runs polling and public opinion research firm InsiderAdvantage, said some polls are missing a key dynamic. He calls it the “silver wave” — a possible surge of older voters that could expose flaws in the polling and reshape the outcome on Nov. 3.
Early voting offers some hints. In Georgia, fewer than 70,000 early votes have come from 18- to 29-year-olds, compared with more than 549,000 from voters 65 and older. Ohio tells a similar story: 406,000 early votes from the white-haired crowd, against fewer than 50,000 from the 34-and-under crowd.
“All of this nonsense you’ve heard about young people storming the polls and they’re going to change everything — it’s not taking place,” Mr. Towery said on Fox News.
Robert Cahaly, who runs the polling and market research firm Trafalgar Group, said voters 50 and older are far more willing to vote than younger Americans, who often tell pollsters they’ll cast a ballot and then don’t follow through.
He also said seniors who were less enthusiastic about President Trump in 2024 than about past Republican nominees are drifting back toward the GOP.
“A lot of the seniors are coming back, and they might not all love Trump,” Mr. Cahaly said.
He attributed the shift to older voters being more supportive of President Trump’s approach to Iran, citing memories of the hostage crisis, and being less affected by high gas prices because they drive less, if at all, and are better positioned financially.
Older voters are also worried about the rise of socialism in the U.S., he said.
The bigger problem, Mr. Cahaly said, is that the polling industry’s traditional tools just don’t work like they used to.
Automated phone surveys have been hobbled by the disappearance of landlines, and voters are increasingly unwilling to answer calls from unknown numbers or sit through long online surveys.
He also knocked polls that ask 20 or 30 questions, saying they tend to attract highly engaged voters and people with time on their hands.
“How many people do you know that are under 50 that are going to answer the phone verbally from a number they don’t know?” he said.
Mr. Cahaly said many media and university polls overrepresent college-educated and higher-income voters, who have trended Democratic and have more time to kill, while missing the working-class and older voters who have moved toward Republicans.
He’s not alone in questioning whether the polls capture the electorate.
Frank Luntz, a longtime pollster, said the industry’s biggest challenge isn’t measuring public opinion. It’s figuring out who will actually cast a ballot.
He said Republicans, especially Trump supporters, are less likely to take surveys, and that many polls don’t ask the questions needed to gauge engagement and turnout. Polls, he said, may be missing less-engaged voters who won’t answer a survey but still show up on Election Day.
“I’m not convinced that they’re polling the people actually going to participate,” he said. “Turnout determines everything.”
Similar doubts have popped up elsewhere in the polling world.
In a recent interview with Politico, Republican pollster and strategist Patrick Ruffini said he thinks many surveys are right that the environment favors Democrats but may be overstating how big the battleground is.
Mr. Ruffini pointed to states like Ohio, Iowa and Kansas, where he said history suggests Democrats face steeper odds than recent polling implies.
He also said pollsters are wrestling with response-rate problems that appear to favor Democrats.
“When you’re seeing high-propensity Democrats at twice the rate of high-propensity Republicans, when we know that the turnout on both sides would be roughly evenly paired, that’s a warning sign for polling in this cycle,” he told the news outlet.
Others have questioned specific surveys.
Take a recent University of Texas poll showing Democratic state Rep. James Talarico leading Republican Attorney General Ken Paxton by 3 percentage points in Texas’ U.S. Senate race, even though the Republican got just 55% of the White vote in the survey.
“Even Ted Cruz had 66% of the white vote in 2018,” said former Army Capt. Seth Keshel, an election analyst. “So those kinds of things you look at it, it’s like … either all these white rural voters that have voted Republican for everybody are going to run 20 points left, or this poll is going to be dead wrong.”
Look back at 2024: Mr. Trump carried Kansas by 16 points, Texas by nearly 14 and Ohio by 11. He also won Iowa by 13 points after the final Des Moines Register poll showed Kamala Harris up by 3 points.
Others cite GOP Sen. Susan Collins of Maine, who was left for dead in 2020 after polls showed her losing by 6 points. She won by 8 points. She’s trying to repeat that this year against Democrat Troy Jackson, who is up by almost 2 points in the latest Real Clear Politics polling average of recent polls.
Mr. Keshel also questioned prediction markets and polls that suggest Democrats are competitive in deeply Republican states.
He said that Mr. Talarico may become the first Democrat to win statewide in Texas since 1994, but it would take a near-perfect alignment of depressed Republican turnout, crossover voting and overwhelming Democratic margins in urban areas.
He applied the same skepticism to Kansas, where Democrats have struggled statewide for decades.
“My beef with the whole system is I don’t think that pollsters are considering any reality,” he said.
Please read our comment policy before commenting.