We’ve heard a lot of talk about how the 2016 and 2020 polling underestimated support for Donald Trump. In 2020, for example, polls showed Joe Biden crushing President Trump. Before the election, Biden led by 8.4 points in FiveThirtyEight’s average and 7.2 points in RealClearPolitics’ average but won the popular vote by less than 4.5 points—enough for an Electoral College victory. Moreover, Biden’s predicted dominance in Pennsylvania, Wisconsin, and Nevada turned out to be more like nail-biters than emphatic wins.
The American Association for Public Opinion Research called this the largest polling miss in 40 years, as polls overestimated Biden’s final advantage.
Scott Keeter, senior fellow at the Pew Research Center, told Newsweek this week that polls in 2020 and 2016 were inaccurate because Trump supporters opted out of surveys due to distrust in institutions like the mainstream media and polling organizations.
“Trump supporters may have less trust in the institutions sponsoring these polls. If they choose not to participate because of this distrust, their support isn’t accurately captured,” Keeter said.
But this year, experts believe polling will be more accurate this year due to adjustments for previously underrepresented Trump supporters. Polling methods have evolved significantly since 2016 and 2020. Back then, many surveys still used landlines, even though fewer than half of U. S. households had them. By 2022, 61% of U. S. polling organizations that conducted national surveys in 2016 had updated their methods.
“Here’s the bottom line,” said CNN polling analyst Harry Enten. “I think a lot of folk are counting in that Donald Trump will in fact be underestimated by the polls, but when I’m looking at the evidence, I think that there are folks who are underestimating the idea that maybe Kamala Harris will be underestimated by the polls, at least a week out.”
PoliticsRSS Fact: No party has been underestimated in swing states for three consecutive presidential elections since at least 1972.
Are the polls underestimating Harris’s support this year much like they did Trump’s in 2016 and 2020?
If the 2024 polls resemble 2022 polling, Kamala Harris could win in a landslide. Two years ago, the midterm polls underestimated Democrats by 4 points, leading them to win 51 Senate seats while Republicans barely took the House.
The expected red wave was more like a red drop.
If this pattern repeats, Harris might secure 319 electoral votes to Trump’s 219, winning all crucial battleground states and the presidency.
Polls suggest Harris is leading the popular vote, while Trump may win the Electoral College. Nate Silver’s forecast shows Harris 0. 9 points ahead with a 71% chance of winning the popular vote, but only a 44% chance for the Electoral College compared to Trump’s 55%. FiveThirtyEight shows Harris 1. 4 points ahead nationally, yet gives Trump a 52% chance of winning compared to Harris’ 48%.
Aggregators show Trump leading in Arizona, North Carolina, Georgia, and Pennsylvania, while Harris leads in Michigan and Wisconsin. FiveThirtyEight shows Harris 0. 1 points ahead in Nevada; Silver’s forecast has them tied.
Given the razor-thin margins between candidates in several states, even a small error could change the winner. Consider Georgia, it’s Trump’s most favorable swing state and he currently leads there by 2.6 percentage points according to the Real Clear Politics aggregate of top pollsters.
If there’s a repeat of the underestimation for Democrats we saw in 2022, Harris could win Georgia by 1.4 percentage points… And thus far, the polls seem to be underestimating the level of female voter participation – and women’s support for Vice President Harris.
If the polls are inaccurate again this year, there’s a strong possibility they might skew more Republican rather than Democrat. If so, it will raise an important question—just how significant will this discrepancy be?
