Will Trump Outperform Polls Again in 2024?

Until the final election results are in on Nov. 5, or possibly later if the race is close, we won’t know how accurately the polls predicted the outcome. However, as the election nears, three main possibilities emerge based on current polling and trends from previous elections.

One possibility is that Trump is leading by more in swing states than current polls predict. This theory is based on the fact that in 2016 and 2020, Trump overperformed in the election compared to the polls.

This theory is supported by polling data from 2016 and 2020. On average, in the seven 2024 swing states, polls underestimated Trump’s 2016 performance by 1.8 points compared to the final election results. In 2020, the polls were more accurate, with only a 1.1-point bias in Biden’s favor. Most of that difference occurred in Wisconsin, where the RCP Average had Biden up 6.7 points, but he won by only 0.7. This effect was also seen nationally, as Trump did 1.1 points better than predicted by the 2016 RCP Average and 2.7 points better than the 2020 RCP Average.

If this pattern continues in 2024, Trump will win, as he currently leads in every swing state by an average of one point in the RCP Average of the Battleground States. With an extra 1-1.5 points added, he would win comfortably.

However, Trump outperforming the polls is not a guarantee, as an alternate hypothesis points to the 2022 midterms, where the opposite occurred. Polls at the time predicted a “red wave” that never materialized, as Republicans failed to win back the Senate and only narrowly took control of the House.

In six of the seven 2024 swing states, there were Senate elections in 2022, and unlike in 2020 and 2016, the polls actually showed a pro-Republican bias; on average, the polls overestimated Republican candidate performance by 3.7 points. The generic congressional ballot was more accurate, with the RCP Average only 0.3 points off from the final results.

Those who believe the polls will continue to be biased against Trump point to the fact that 2022 was a midterm election, with different voter demographics than in presidential elections. In 2020, 154.6 million people voted, compared to only 107.7 million in 2022. Since Trump wasn’t a candidate in the 2022 midterms, pollsters may rely more on their methods from past presidential elections, potentially leading to a bias in favor of Democrats again.

Supporters of the idea that Trump will underperform the polls cite the most recent midterm election in 2022, suggesting pollsters might overcompensate for their previous anti-Trump biases from 2016 and 2020. If that happens in 2024, Trump’s current 1-point lead in the swing states could be undone.

A third possibility is that pollsters have learned from previous elections, and there won’t be much bias in either direction. This could be the case, as this is Trump’s third time running for president, giving pollsters more data on voting patterns and whom the likely voters are.

If the polls don’t show bias in either direction, Trump is likely to win, but it’s not guaranteed. Polls could still be off in swing states, especially considering that the RCP Averages in Nevada, Wisconsin, Pennsylvania, and North Carolina all have the race within a single point.