All news
US presidential election voting nears completion: results could be problematic
It may seem counterintuitive, but we are actually within the confines of the typical electoral pitfalls that separate any candidate from a decisive victory, especially in the Electoral College, writes The New York Times.
A one-sided result, in which only a small margin is expected, could further increase distrust in the vote count and the electoral process itself.
"The popular vote is very close, and it's possible that someone could overturn the 315 Electoral College votes, but it doesn't seem that far-fetched," said Lee Miringoff, director of the Marist College Institute of Public Opinion.
"Or they could win the popular vote by five points," he added. "By today's standards, that's a stunning result. "That's a word no one has used this year."
Since 1998, the margins of votes in presidential, House, Senate, and gubernatorial races have been within 6 percentage points of the final vote, on average, according to a FiveThirtyEight analysis. But in the 2022 midterms, the average margin of error was 4.8 points, making it the most accurate poll in 25 years. If the votes were tied and the votes were evenly split this year, the winning candidate would win decisively.
According to a polling average released Monday, if the polls underestimate Kamala Harris by 4.8 points in each of the seven swing states, she will win every state and receive a total of 319 electoral votes, while Donald Trump will receive 219 electoral votes. Trump will win every battleground state, receiving a total of 312 electoral votes by an equal margin. (These calculations assume that Harris and Trump win other states they like, the New York Times points out.)
Poll results may differ in one direction or to different degrees. But even a historically accurate election year could mean that either candidate wins a majority of states, including Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania, and Wisconsin.
Election polls have their share of errors. Partly because of the voting process itself, but mostly because they have to guess who will actually vote in order to match their sample to that group of voters. Other polls, on non-political topics, should not touch on this question.
"Because the tools that pollsters use to predict turnout (stated voting intentions, interest in the election, perceived importance of the outcome, historical turnout) are so limited, if those tools don't prove useful in an election, all you can do is make the same mistake," said Scott Keeter, a senior polling consultant at the Pew Research Center.
Obviously, who comes out to vote will determine who wins. Right now, Trump’s biggest swing state advantage is Arizona, where the polling average shows Trump leading by 3 percentage points. So even a small drop of 2-3 percentage points could result in a big win for both candidates.
This is partly because polls tend to show the same results over a given election cycle (though not always), and battleground states with similar demographics tend to vote the same way (though not always). As a result, if polls show even small differences in the same “direction” (under-Trump or under-Harris), it can mean the difference between one candidate or the other, which can lead to racial dominoes.
None of this necessarily means there will be a knockout on Election Day. The New York Times notes that there is still a very real possibility that this will be a close election, with the election coming down to just a few hundred votes in a single state. But knowing that these possibilities are within the typical margin of error of polls could help mitigate some of the impact when the final results are tallied.