General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsPLEASE NOTE: You cannot compare 2020 'exit poll' results with 2016
There is no direct comparison between
White Women 2020 v 2016
Black Men 2020 v 2016
Latinx 2020 v 2016
The consortium took the election day questionnaire results, then tried to adjust for the 100 million early voters with telephone polling (and scattered questionnaire results from some early voting sites ). The election day results were skewed GOP but there was no independent data to unwind the overall bias, let alone individual biases amongst the demographic groups. So while the results are not totally unreliable, they cannot, CANNOT be compared with the exit questionnaire results from 2016 and earlier.
So you cannot say, for example, that Biden scored less well with black male voters, by comparing the 2020 'exits' with the 2016 numbers. This may or may not be true, but you are not comparing like with like.
Be Wary Of Exit Polls This Year
https://fivethirtyeight.com/videos/be-wary-of-exit-polls-this-year-well-and-all-years/
Ferrets are Cool
(21,106 posts)pwb
(11,265 posts)Trumps have made lying the norm. People will lie about poles for political purposes Right? Joe won by what 4% so that is close within the margins of error.
Tribetime
(4,696 posts)Seems every poll was off in favor of Trump 46 of 50 states outside moe...4. Within 1 % Georgia Colorado D.C. and I think Oregon
pat_k
(9,313 posts)And back then, it was a powerful tool to detect vote "subtraction." One that is used by election monitors around the world. 2001 and 2004 exit polls in many states was spot on with reality -- pre-suppression reality. As investigations like Conyer's of Ohio in 2004 showed.
Still Sensible
(2,870 posts)"other than in-person, polling place voting" renders exit polling rather worthless. It doesn't matter if pollsters tried to "fix" the problem with other supplemental means.