Could polls about battleground Senate races be wrong? Democrat, Republican strategists weigh in

Democrat and Republican strategists weigh in on whether polls in Pennsylvania, Ohio, Georgia and Minnesota could be wrong in predicting Dem candidates winning in November

An eventful summer of abortion fights, bruising GOP primaries and a rise in President Biden's approval rating have turned near certainty that Republicans retake the Senate into a doubtful prospect.

Polls in battleground Senate races across the U.S. in recent weeks have shown GOP candidates behind their Democratic rivals in public opinion polls, with many citing the Supreme Court's overturn of Roe v. Wade as a motivating issue for the left, as well as independent swing voters.

But polls have been wrong before. In 2016, numerous polls showed Hillary Clinton ahead of Donald Trump, who succeeded in winning the election. Polls got major Senate races wrong in 2018 and 2020 as well – including Sen. Susan Collins, R-Maine, who won re-election in 2020 by 8.6% despite most polls showing her underwater. 

To understand whether 2022 polls could have the same issues, Fox News Digital asked several top strategists, from both the Democratic and Republican parties, if the current 2022 polls that indicate Democrat Senate candidates are leading in states such as Pennsylvania, Ohio, Georgia and Minnesota could be as wrong as opinion polls were in 2020 and 2016. Here's what they said.

SEVERAL GOP SENATE CANDIDATES REVEAL THEY DO NOT SUPPORT 15-WEEK FEDERAL ABORTION LIMITS

"It’s quite possible the polls are off again because Democrats are performing better than we might expect given past elections in the same places they were off in 2020 and 2016. But more – pollsters haven’t changed the way the poll because AAPOR (polling community) couldn’t agree what went wrong in 2020 and thus didn’t significantly update their methodologies."

"Some people never learn. The pollsters who blew it in 2016, 2018, and yes, in 2020, are at it again, predicting gloom and doom for Republican Senate candidates and pretending a deeply unpopular president and vice president are surging'. If past is prologue, these polls overstate strength of the Democrats and undercount right-leaning Independents and Republicans. 

"With a slight change following their embarrassing 2016 polling miss, they missed again in 2018 and 2020. In 2018, they wrongly predicted the Republicans would lose the Senate and that Democrat Sens, Bill Nelson (Florida) and Joe Donnelly (Indiana) would win re-election. The opposite happened."

"Even more glaring, in 2020, not a single poll in the Real Clear Politics average showed Sen. Susan Collins of Maine winning; she triumphed by 8.6%. In North Carolina, not a single public poll in the closing weeks showed Sen. Thom Tillis winning, yet he did. In Iowa, Montana and South Carolina, public polls went back and forth showing a true toss-up. Yet the actual results were hardly a squeaker. Sen. Joni Ernst won by 6.6%, Sen. Steve Daines by 10% and Sen. Lindsey Graham, facing a Democrat opponent who raised $112 million, beat him by 10%.

"Keep these facts and phony figures in mind when you see polls in swing states like Arizona, Pennsylvania, Ohio, Georgia and Nevada showing comfortable leads for tax-and-spend-happy candidates that allow inflation to spike and criminals to roam free and yet trap our kids in failing schools. 

"Don't fall for it. Voters don't ask, ‘Who can win?’; they ask, ‘Who can lead?’ These candidates must ignore the critics and the phony polls, velcro Democrat candidates to Joe Biden and his failing and flailing, and give voters a reason to believe, trust and support them."

"The polls were wrong because they underestimated the proportion of working-class voters. And Trump brought out tens of millions of new, first-time voters. But it was there in the registration, all year in 2020. I assume all polls, like mine, use the 2020 non-college proportion as given." 

"If the same polling error was to repeat itself, it would be evident in the new registration. I think women are dominating the new registration."

WASHINGTON AND OREGON VOTERS SAY ABORTION, AFFORDABLE HOUSING AMONG TOP ELECTION PRIORITIES

"Yes, the polls could be as wrong as they were in 2016 and 2020. We know that in those cycles they undercounted Republican voters, perhaps because Trump voters do not trust or like pollsters and/or the media. But in both of those elections, Trump was on the ballot and turnout among lower-propensity voters was much higher than we’re likely to see in a midterm. Those are the voters most likely to be undercounted. That means the better comparison is probably to 2018, when Trump wasn’t up, such voters did not turn out, and the polls were much more accurate."

"That said, Democrats must presume that they polls are wrong and that they are doing much worse than the polls would indicate. Running as if you’re behind is the only proper course in elections that are likely to be very close." 

"There is no doubt that the polls could be wrong – but It's important to understand some of the reasons why, and to also make the distinction between a poorly executed public poll, and a more sophisticated voter poll that campaigns use for their strategy and messaging and that the public rarely gets to see.

"Some of the polls are done on the cheap by people who are more interested in headlines or ratings than in methodology. So you get polls that are skewed, and when you try to find the breakdown of the sample, it's almost impossible, or it's way off.

"Some polls are weaponized. They are released to further a particular narrative, and discourage some voters from participating in the election, because they think the race is a foregone conclusion.

"What does a poll of likely voters even mean? Is it a random sample of adults who SAY they are likely to vote, or is a poll taken from a list of people who actually vote in midterms and have the vote history to show their participation? These are important questions because they can determine whether a poll is skewed – or just plain wrong. But oftentimes, this information is hard or impossible to assess.

"Sometimes a poll can be right, but the analysis is wrong. For example, a poll that shows two candidates with 40% each might be reported as a ‘tie score,’ but the internal cross tabs might show that the undecided voters will break 3:1 to one of the candidates, so the real outcome will be 55%-45%. But that is overlooked, and the only thing reported is te 40%-40% tie. The poll showed what would happen, but it was overlooked or under-analyzed, and perceived as wrong."

Christy Setzer, Democratic strategist and founder of New Heights Communications

"It's always possible that the polls are wrong – at this point, we expect them to be. But two factors are present this year that play in Democrats' favor: the lack of Trump on the ballot, and the huge, possibly unprecedented swell of voter enthusiasm from younger women who are fired up post-Dobbs decision to vote against Republicans' anti-abortion agenda.

ABORTION REMAINS KEY ISSUE FOR SOME VOTERS AHEAD OF MIDTERMS

"You can see a changed electorate coming to the polls since the Dobbs decision, resulting in Alaska's first Democratic representative in 50 years, in Kansas overwhelmingly rejecting the anti-abortion ballot initiative, and in Democrats winning or being competitive in places they frankly shouldn't be. We should expect some surprises on election night, given that." 

"The polls might be wrong but any candidate that looks at polls any day before the election is a fool. Sen. Manchin always told me you run every race like you are behind. 

"A more important indicator of election results is the quality of the Democratic candidates, the extreme views of their opponents and support from voters on key issues. 

"If Democrats run on their historic legislative successes and protecting women’s health care and Republicans run on Donald Trump’s list of grievances, it’s going to be a great night for Majority Leader Schumer and the growing Democratic caucus."

Data & News supplied by www.cloudquote.io
Stock quotes supplied by Barchart
Quotes delayed at least 20 minutes.
By accessing this page, you agree to the following
Privacy Policy and Terms and Conditions.