Whether people are willing to tell strangers their vaccine status over the phone is actually one of the things that we have an excellent way of testing -- there's CDC data to compare it to!
So far, many public polls tracking this look pretty accurate.
None of this is to say that election polling doesn't have ongoing issues, or even that polling around vaccines/mandates isn't subject to them (there are some less rosy assessments out there as well). But this is really one of the *more* directly measurable questions out there.
no better way to realize that you are an Old than the sinking feeling of looking through 2008 polling and realizing that it really was an entirely different political era
Also, this: "[T]he problem isn’t really polling. Rather, it’s the use of polls to center everything that happens in politics around those November 2022 swing voters in Wisconsin."
The subset of political actions that are going to meaningfully affect an election is both small and often unknowable, and yet that often becomes the primary lens through which those actions are examined, even in cases where there are other large, knowable effects.
One of the reasons election polls get so much attention (beyond the obvious) is that they serve as an external validator of polling data. Here, the CDC numbers allow us to compare the polls with a source of hard data on the actual number of shots.
(Another interesting outcome of having that data is that a couple of pollsters have recently started weighting *to* the CDC vaccination rate as a benchmark -- will be very interested to hear more on how that's playing out.)