Polls have been crapping out lately usually hiding Republican Votes. Maybe Dem leaning voters have stopped answering also?
Maybe polls are now just dated and shouldn’t be trusted.
Well, if you (or anyone) happens to run across some kind of post-mortem, I’m interested.
For what it’s worth, some of the bigger liberal Substackers (such as David Kuo and Simon Rosenberg) kept showing NJ Gubernatorial polls that consistently had Sherrill up 7-15 points. They tended to treat any poll showing a close race as, essentially, BS polling gamed by the GOP.
Turnout was practically record-breaking for an off-year, so maybe that screwed up the pollsters models?
I have no evidence, but I think it’s likely this: some pollsters over-corrected to make up for an expected gap in phone survey responses between the parties, but now pretty much everyone hates answering a stranger’s phone call.
Could be.
Polling averages were off in one direction in New Jersey/Virginia and wrong in the other direction in New York City.
State and local polling just isn’t as accurate as national.
This is somewhat of one:
This one is paywalled, but if you have access to WSJ, they talk more about the pollsters in this article.
I don’t have access to the WSJ, unfortunately.
But, yeah, that swing in the Hispanic vote may have been underestimated by the pollsters, too.