I've wondered specifically about this automatic-spam-call-blocker aspect of The Polling Issue, and whether we're like, sampling only people who have old or shitty cell phones or something?
I'm a millennial but my phone has a spam blocker that doesn't even show me the calls that would be coming from a pollster. And NYT/Siena is all phone. You'd have a way better chance of reaching me via the Pew method of sending out physical mail asking you to participate online. I get mail!
We are sampling elderly people with land lines who answer calls from people they don't know.
Which is why polling absolutely blows. And why 'polls say Biden behind by 3 points'. That means nearly half of the traditional GOP base favors Biden.
And why I ignore polls.
How do they get different age demographics in the crosstabs then? I am not a polling enthusiast but they are a little more sophisticated than this stereotype.
I'm assuming that whatever young people they do reach must be biased in some I'm assuming that whatever young people they do reach must be biased in some way. Maybe they live with an elderly person? That could indicate living at home or assisting an elderly relative with Healthcare
My understanding is that most companies these days do call cellphones etc., but apparently only 1 person in ~100 picks up and agrees to participate, so they're probably oversampling certain types of people.
I realize my comment is a sweeping generalization. But it simplifies things to illustrate that polling methods haven't changed as our phone usage has. Pollsters are scrambling to find new ways to sample people. But the most accurate poll is still an exit poll. Which is why 1/
2/ people are repeatedly astonished that election results are so often wildly different from polling results. 'But the polls said . . . ' When we see polling results reflected in election results, then I'll begin to trust polls again.
