Second poll that messes with historical demographics to show Harris barely ahead
Survey USA has a new poll out that shows Harris winning the state by one point in spite of losing Independents by four. This is basically the same issue we saw with the Quinnipiac poll showing Harris ahead in spite of favorable demographic numbers for Trump.
So the Demographics in North Carolina in 2020 were 37% GOP, 34% Dem, and 30% IND. In 2022 the GOP increased that three point advantage to a five point advantage in voter turnout (37-32-31). Both the Quinnipiac and SUSA polls are suggesting that the Democrats will be holding a voter turnout advantage of somewhere between 1-3 points to get to their results showing Harris winning. Why do they come to this conclusion? No clue. North Carolina is one of the states where Republicans have managed to out-register Democrats so not sure why any pollster would assume that Democrats will suddenly win the turnout battle, especially considering the slide the Democrats have had over the past couple of elections in that state.
Now the SUSA polls shows Trump winning the GOP voters by the same amount as Harris does with Democratic voters, and it shows Trump leading with Independents by four points. If you plug in the 2020 turnout and model it... you get Trump up by nearly four points. According to this, the poll is overstating Harris support by nearly five points.
To be clear, in 2020 Trump won that state by 75,000 votes while losing Independents by four points. In 2022, Republican Ted Bund lost independents by a similar margin in North Carolina and still pulled off a victory by over 120,000 votes. Are we to believe that Trump now pulls in 93% of his GOP vote, has a lead with Independents, but will still lose? The concept is ridiculous.
For the most part the polling I have seen in most of these states has been at least reasonably believable. But these last two polls showing Harris leading in North Carolina are not reasonable at all.
Funny... I went to that X link and saw a post from Indy Voter! Anyone remember him?-
~~datahazard~~ on X: "Only 53 voters at one address? This Philadelphia Homeless Servicing NGO has 5,200 voters at their one address. https://t.co/Z8kPqTtixW" / X
Truth in advertising -
Nate Silver:
Last update: 2:45 p.m., Tuesday, October 22. Strong day of state polling for Trump, with leads in high-quality polls of Georgia and Nevada — where early voting numbers also look promising for Republicans. Trump has steadily improved in our forecast over the past two weeks, but the model is mostly sticking to the view that these are incremental shifts rather than a sea change and the race remains highly competitive.
I wonder how long the incremental shift has to continue before Silver acknowledges that this race should no longer be considered a toss up? It seems like a small Harris lead in his model was worth quite a bit more than a similar Trump lead.