Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News Editorials & Other Articles General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

RandySF

(66,226 posts)
Fri Sep 6, 2024, 06:02 PM Sep 6

Democratic operative explains what caused the 2016 polling error

On the podcast 'The Downballot', Democratic data expert John Hagner explained that Trump voters did indeed share their voting intentions, but hung up before completing the survey, resulting in incomplete calls that did not show up in the data and skewed results.

10 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
Democratic operative explains what caused the 2016 polling error (Original Post) RandySF Sep 6 OP
Ahhhh...aren't those incomplete results not to be included? brush Sep 6 #1
Correct RandySF Sep 6 #2
No patience senseandsensibility Sep 6 #3
I should note. RandySF Sep 6 #4
Why would they do that and how does he know that bucolic_frolic Sep 6 #5
He's worked for Dems on data since the Howard Dean campaign. RandySF Sep 6 #6
It implies that pollsters track how many hang up and shared that metric with him. bucolic_frolic Sep 6 #10
It's a myth that polling was all that far off in 2016 anyway RidinWithHarris Sep 6 #7
Why is anyone concerned about an obsolete poll? Think. Again. Sep 6 #8
What about in 2020? democrattotheend Sep 6 #9

RandySF

(66,226 posts)
2. Correct
Fri Sep 6, 2024, 06:13 PM
Sep 6

If a client pays for a 15-question poll, the pollster is obligated a 15-question interview. So, is someone says "Because fuck you" and hangs up, the couldn't use the candidate preference data. Hagner said about a third of Trump respondents did this.

bucolic_frolic

(45,767 posts)
10. It implies that pollsters track how many hang up and shared that metric with him.
Fri Sep 6, 2024, 06:35 PM
Sep 6

And that Republicans did it more than Democrats. Interesting.

RidinWithHarris

(381 posts)
7. It's a myth that polling was all that far off in 2016 anyway
Fri Sep 6, 2024, 06:23 PM
Sep 6

538 had Clinton's chance of winning at 70%. That's nice, but hardly a shoe-in. The 30% possibility coming true is hardly that damning of the polling in a tight race.

It was people's expectations and punditry that were way off base, much further off base than the raw numbers.

Think. Again.

(15,405 posts)
8. Why is anyone concerned about an obsolete poll?
Fri Sep 6, 2024, 06:35 PM
Sep 6

Polls are not science, they are shots in the dark OR creatively crafted narrative starters.

Oh, wait, I get it, by pretending the polls from the past were wrong for...reasons, we can continue to pretend current polls are perfectly correct now. Got it.

democrattotheend

(12,006 posts)
9. What about in 2020?
Fri Sep 6, 2024, 06:35 PM
Sep 6

Did they include those calls? If so, that could explain why the downballot Dems underperformed, if they didn't stay on long enough to answer the downballot questions. That could also explain why so many downballot Dems are outperforming the top of the ticket this time.

Latest Discussions»General Discussion»Democratic operative expl...