Sep 26, 2026 · 45 min
Polling skepticism meets a stronger Democratic landscape
Can We Actually Trust the Midterm Polls?
The episode separates genuine weaknesses in polling from reputational damage left by past misses, with consequences for how voters read the midterms.
- 1Democrats may hold an unusually strong position as Donald Trump’s approval declines and broader electoral conditions shift.
- 2Past polling misses deserve scrutiny, but the 2018 and 2022 polls were better than their reputations suggest.
- 3Primary polls, weak respondent verification, loaded wording, and online political noise can all distort the picture.
Don't miss
Jain details how unverified polling respondents, including bots and fake residents, can undermine surveys before the results are modeled.
The brief
Peter Hamby and Lakshya Jain begin with an unfashionable premise: Democrats may be in an unusually strong midterm position, despite widespread distrust of polling.
Jain argues that analysts overcorrect after perceived polling failures, noting that the 2018 and 2022 polls were better than their popular reputations.
The conversation weighs campaign whispers and political gut feelings against quantitative models, with Jain arguing that meaningful effects should appear in the data.
Economic dissatisfaction helps explain voters’ movement from Biden toward Trump, while low response rates and partisan spin make motives harder to measure.
The sharpest warning concerns poll quality: unverified respondents, fake residents, bots, and loaded questions can make credible-looking surveys misleading.
Hamby and Jain close by separating political social media from ordinary voters, especially swing voters less immersed in partisan online discourse.