Opinion polls underestimated Donald Trump’s support level for the third consecutive US presidential election, predicting a tough showdown with Kamala Harris after the Republican finally overtook the vice president in battleground states. Trump’s victory included growing support across many demographics and regions, but experts said pollsters failed to accurately predict the race in states where results differed significantly from previous elections in 2020.
“They performed well on the battlefield, but … they failed to provide the essential information that Trump was pushing on board,” said Michael Bailey, a political science professor at Georgetown University.
According to The New York Times, more than 90 percent of US counties voted in higher numbers for the Republican billionaire than in 2020.
Overall, polls had predicted little difference in the race in the seven battleground states that decide US elections. As of Wednesday, Trump was projected to win five of those states by between one and three percentage points.
According to those projections, the former president was on track to win all seven states.
“Trump may have been underestimated a little bit, but I think the polls did pretty well collectively — it wasn’t a big mistake,” said Kyle Kondik, a political analyst at the University of Virginia.
“The polls showed Trump had a good chance of winning, and he won.”
Pollsters’ performance this year was under the microscope, after two major misses in a row: They failed to anticipate Trump’s victory in 2016, and the margin by which President Joe Biden won against him in 2020. Was underestimated.
“Trump is underreported by about two points this time in key states,” said Pedro Azevedo, head of US polling at Atlassian.
In Pennsylvania, the latest polling average from RealClearPolitics gives Republicans a lead of 0.4 percentage points. Till Wednesday he was ahead by two points.
In North Carolina, polls predicted a 1.2-point margin for Trump, and he won over Harris by three points.
In Wisconsin, the Vice President was given a lead of 0.4 points, but projected results showed Trump leading by 0.9 points.
The core problem has not changed since Trump’s arrival on the American political scene nearly a decade ago: A swath of his voters refuse to participate in opinion polls, and companies have failed to be able to accurately estimate his impact. .
NYT data analyst and polling guru Nate Cohn wrote two days before the election that, in the most recent poll conducted by The New York Times with Siena College, “white Democrats were 16 percent more willing to respond than white Republicans.”
He said this disparity has increased during the 2024 campaign.
Although pollsters such as The New York Times/Siena tried to compensate for these imperfections with statistical adjustments, it was clearly not enough.
“It’s clear that the polls have significantly underestimated Trump’s lead among Hispanic voters,” Azevedo said, pointing to Trump’s larger-than-expected wins in Nevada and Florida.
“The situation is similar among white voters,” he said, adding that while most polls expected Harris to “improve her margin” in this demographic, Trump fared better in the polls and increased her numbers in rural areas. Took.
Iowa was a prime example of this, where a poll three days before Election Day found Harris with a three-point victory in the solidly Republican state. In the end, Trump won it easily by 13 points, Azevedo said.
The author of that inaccurate Iowa survey, J. Ann Selzer said the difference could be made by voters deciding late.
“Late decision makers could have chosen Trump in the final days of the campaign after the interviews were completed,” he told the Des Moines Register newspaper.
“People who had already voted but chose not to tell our interviewers who they voted for could have given Trump the edge.”
(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)
