As 2024 looms, the battleground states -- Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania, and Wisconsin -- are once again at the heart of the contest. Current polling shows Trump leading in some of these states, with others leaning toward his opponent, Kamala Harris, but the margins are slim across the board.

The question is: what happens if polling errors similar to those in 2016 and 2020 persist in these states? By applying a weighted average of the polling discrepancies from the last two election cycles (with more weight given to the more recent 2020 election), we can estimate how much Trump’s support may be underrepresented in 2024.

Here are the state-by-state adjustments based on historical polling errors:

• Arizona: Polling underestimated Trump by an average of 2.8 points.
• Georgia: Trump was under-polled by 1.8 points.
• Michigan: Polling was off by 4.4 points.
• Nevada: Nevada has been more accurate, with an average error of only 0.4 points.
• Pennsylvania: Trump was under-polled by 4 points.
• North Carolina: Polling errors averaged 2.6 points in Trump’s favor.
• Wisconsin: Polling underestimated Trump by an average of 7.4 points.

Applying the 2016 and 2020 Errors to 2024

Now, let’s apply these historical polling errors to the latest 2024 polls. Current data shows Trump leading in Arizona, Georgia, and North Carolina by narrow margins. Meanwhile, Harris is slightly ahead in Michigan, Pennsylvania, and Wisconsin, with Nevada a virtual tie.

However, when factoring in historical polling errors, these margins shift significantly. In Michigan, where Harris currently holds a 1.8-point lead, applying the average 4.4-point underestimation of Trump flips the state in his favor. In Wisconsin, Harris’s narrow 1-point lead is reversed by the 7.4-point polling error, giving Trump a substantial lead. The same dynamic holds true in Pennsylvania, where Trump’s underpolled support could move him from a slight deficit to a solid advantage.

If these polling errors hold, Trump could win seven out of eight battleground states. Winning all reliably Republican-leaning states plus these battlegrounds would comfortably deliver Trump over 300 electoral votes, far more than the 270 needed to secure the presidency.

Is This Outcome Likely?

If past polling errors repeat, Trump is positioned for a strong electoral outcome. Pollsters have certainly attempted to adjust their models since 2016 and 2020, but it remains to be seen whether they have adequately accounted for the factors that led to these systematic underestimations of Trump’s support. Factors such as "shy" Trump voters, undersampling of rural voters, or incorrect turnout assumptions could once again lead to polls that underestimate his true level of support.

Polling errors are not new, nor are they exclusive to one side of the political spectrum. However, the magnitude of these errors in two consecutive election cycles when Trump was on the ballot is noteworthy. Given the closeness of many battleground states, even a small polling error could shift the outcome.

Conclusion: A Data-Driven Caution

This analysis does not suggest that Trump will win in 2024, but it does show that polling data -- if uncorrected for past errors -- may be misleading. It serves as a reminder to take pre-election polls with caution, particularly in close races where small margins of error can have outsized impacts. With a historical pattern of underpolling Trump in several key battleground states, the potential for a repeat scenario in 2024 is something that both pollsters and political observers should seriously consider.

As we approach Election Day, the takeaway is clear: even if polls show a close race, history tells us that the final result could look very different, especially where Donald Trump is concerned.

Image: Defvense Visual Information Distribution Service