Read on Twitter
Well, I'd posit that if the Trump voters were lying, that we'd see a lot of GOP primary voters / reg. GOP saying they were voting for Biden (as we know the reg/vt hist of our respondents; not subject to lying). But we don't see that. https://twitter.com/eweingart02/status/1317988751176617990?s=20
Well, you know the party registration of absentees. Our poll reports their stated pres. vote preference. Those aren't the same.
For ex: our FL absentee-requesting voters were D+14 by reg., D45-R31.
Those same voters were Biden 61, Trump 30. https://twitter.com/DDTGlobal1/status/1317990521068048390?s=20
For ex: our FL absentee-requesting voters were D+14 by reg., D45-R31.
Those same voters were Biden 61, Trump 30. https://twitter.com/DDTGlobal1/status/1317990521068048390?s=20
Yes, but excelled across the board compared to those who didn't request an absentee ballot from the same party https://twitter.com/Iamjusthere4be1/status/1317992023186604032?s=20
No, but the view I was replying to basically holds something like: 'the polls said biden would win mail voters by, say, 50 but by registration the absentees are, say, D+25, and therefore GOP is beating the polls https://twitter.com/blue_proton/status/1317992025669701634?s=20
And my point here was that our polls help reconcile this--they often show the IRL breakdown of mail voters, but show Biden doing way better. Now, polls could still be wrong, but there's no contradiction https://twitter.com/Nate_Cohn/status/1317991599654174723?s=20
Also as an aside, we do have a respondent who admitted misrepresenting their views in a followup interview because of low initial trust in the survey and ... they were a Dem
To be clear, I don't think this is a big issue. But there's no reason to assume that measurement error only runs only one way
A fine point -- and an issue with the shy Trump analysis in 2016, as well (shy Johnson voters in WI?) https://twitter.com/ericmkingsbury/status/1317993416865714180?s=20
