Presidential Election 2024: Why Were the Polls So Wrong?

This piece is part of a series of articles produced in a special topics class taught by Professor of Law, Jurisprudence, and Social Thought Lawrence Douglas on the 2024 election. Articles may have been reviewed by Douglas as well as other members of the class prior to submission to The Student.

The Monday before Election Day, Nov. 4, most news outlets predicted it would be an extremely close race, leading to major delays in the results. Unexpectedly, any American flicking their TV on at 9 p.m. on Nov. 5 saw that President-elect Donald Trump’s victory was already clear. How did the election results have all the national polls and major news stations mistaken in their predictions? The answer lies in both last-minute voters and Trump’s complicated relationship with the media.

On Amherst’s campus, like other universities and colleges across the country, students knew their ballot choice long before Election Day. As a school with the motto, “Terras Irradient”  (“Let Them Enlighten the Lands”) suggests, many students arrive on campus with an already strong political fervor, which only grows when surrounded by other students who highly value intellectual curiosity. Classes at Amherst promote having an opinion, along with the importance of the voters’ role in democracy, so it is no surprise most students in higher education had their vote decided early on. This is formally known as the “Civic Education Hypothesis” which states that education can provide the skills and knowledge necessary to become an engaged and active voter. However, that is not the case outside of the bubble of higher education. For many Americans, indecision plagued them until the last minute. Trump won the minds of last-minute voters. According to NPR, Trump won last-minute deciders by double digits. Of voters who decided in the last week, 12% more voted for Trump than Harris. This helps explain why the polls were incorrect, since this voting pattern in favor of Trump was too last minute to be picked up by the polls. In fact, polls undervalued Trump’s win in every single swing state — and by multiple points (Trump won Arizona by 5.5%).

Some of the last-minute voters in this year’s election were also previously inactive voters. Inactive voters who rarely show up for even general elections are referred to as “low-propensity voters.” Extracting votes from these low-propensity voters can often be the decider in closely contested races. Well before Election Day, Trump claimed low-propensity votes, and his prediction turned out to be correct.

Trump’s criticism of mainstream media as a weapon against him also helps explain why the polls created by media outlets would undervalue Trump’s success. The reason is not that the media outlets are biased against Trump, but rather that Trump voters are biased against the media. When a Trump voter hears the person they support referring to news outlets as “corrupt”, “evil,” and “an enemy of the American people;” they probably will not be too eager to participate in the polling conducted by those news outlets. This results in an inaccurate testing pool and skewed polling results.

I can’t help but wonder if the inaccurate polls could have fostered false hope for Vice President Kamala Harris and her supporters. False hope in a Harris win could have dissuaded low-propensity voters who were leaning towards Harris from voting at all. If the polls showed that Harris was leading in their particular state, perhaps they figured their vote was not necessary. Granted, it would be very difficult to measure the impact of this theory accurately, but it is still a possibility. For Harris, losing low-propensity voters meant losing the election.