Why did the polls get it wrong about Trump for the third time?
Donald Trump has consistently proved himself as a turnout machine who defies polling. Photo: Getty
For the third election in a row, polls under-estimated Donald Trump’s ability to turn out voters and garner support among Americans.
Trump is leading the popular vote, which Harris had been tipped to win by pollsters, although this could change once all Californian ballots are counted.
Nate Silver, an election analyst, said on his Substack that ultimately the polling was more accurate than in previous elections, but still under-estimated Trump.
“The final margins on Tuesday were actually quite close to the polling averages in the swing states, though less so in blue states,” he said.
“Mostly, it’s deep blue states where Harris under-performed, including her native California.”
This included New Jersey, where Harris was expected to win by 20 per cent and ultimately only won by half of that.
Even in red states, polling averages only had Trump ahead in Florida by around 5 per cent, but he ultimately won the state by 13 per cent.
This isn’t to say that every pollster was inaccurate, with Atlas Intel appearing to come the closest to correctly predicting a win that will ultimately go down in the history books as a landslide for Trump.
They did, in general, come much closer to capturing an accurate reading of the results than in previous years, showing a race that was ultimately won by a thin margin in key states.
Many pollsters introduced new methodologies and herding, where results are compared to the last election to attempt to predict turnout, as a response to the historic misses in previous years.
Big miss
One of the biggest misses of the election was from Ann Selzer, considered to be the gold standard for Iowa polling, whose final poll placed Harris ahead in the traditionally Republican state.
The poll, late in the election cycle, was outside most other pollsters’ range and gave Harris supporters reason to believe she would do well in red states.
The final result was a Trump blowout, he won by 13 per cent, ultimately diminishing the pollster who has accurately gauged her state for decades.
Following the election, Trump’s campaign even took the time to mock the respected pollster when it released a statement that said “President Trump and Vice President JD Vance will help to ease costs, secure the border, and protect Social Security for retirees like Ann Selzer”.
Polling ahead of the 2016 and 2020 elections also under-estimated Trump’s appeal to American voters and helped contribute to the shock of Hillary Clinton’s loss in 2016.
Clinton led Trump in polling for the entire election campaign, before his shock win in 2016. Photo: Getty
Biden’s lead against Trump in 2020 was vastly over-estimated in the polls, with his average lead landing at 7.9 per cent compared to the final result of 4.4 per cent.
Bookies lead the way
Betting markets, however, had Trump ahead as the most likely winner almost the entire way through the contest, resulting in more than one victory lap.
“History was made today,” Shayne Coplan, the chief executive of political betting site Polymarket, wrote on X.
“The Trump campaign HQ literally found out they were winning from Polymarket.”
Others even suggested that betting markets may be a more accurate indicator of results than polling soon, as the United States courts decide whether political betting will be allowed to continue.