In the final stretch before Election Day, the US presidential race between former President Donald Trump and Vice President Kamala Harris is neck-and-neck in key battleground states, according to major polling outlets. Polling aggregator FiveThirtyEight reports an unprecedented 47.8% tie in Pennsylvania, near-even numbers in Nevada, and slim one-point differences in states like Wisconsin, Michigan, and North Carolina.
The razor-thin margins have many wondering if the race is truly this close or if polling accuracy is again in question. American University’s Professor W. Joseph Campbell reflects on how past polling errors in 2016 and 2020, as well as in the 2022 midterms, raise doubts about reliability. “The 2020 election was the worst in forty years for pollsters,” Campbell remarked, noting how Trump’s unexpected support levels consistently skewed previous projections.
Polling experts are working to adjust their methods to better account for Trump voters, but some, like Nate Cohn of the New York Times, caution that these adjustments may now underestimate Harris’s support. Pollsters are grappling with balancing these adjustments, with the added challenge of potential groupthink or “herding” – aligning results closer to consensus rather than risking outlier predictions.
Political analysts Joshua Clinton and John Lapinski suggest this raises a bigger question: are polls reflecting the reality of this election or the assumptions pollsters are making based on past mistakes? As Americans await Election Day, the accuracy of these polls remains under intense scrutiny.