****GUARANTEED FRESHER THAN FRESH QUINNIPIAC FLORIDA POLL**** Biden 51% Trump 38%

[Awsi Dooger]

34. One thing I should add about ideological percentages

They shift in favor of the party with the tide in its favor. Not everywhere but in many states. I have noted that since looking at polls from the '80s. It holds up in midterms and presidential elections. I have never seen a good related study so I've mostly been forced to guesstimate whether the electorate itself changes in the short term, or if the same people merely adjust the way they identify themselves in the short term.

This year if voters are moving away from Trump I would expect 1-2% fewer conservatives in some states and 1-2% more liberals than the norm in some states. That's what happened in 2008. But the problem in states like Texas and Georgia, etc. is that we need a shift of 3% or more, especially the downtick in the conservative category. That is fairly rare. For example, in 2008 only 9 of the 51 states and territories dropped 3% or more in self-identified conservatives from 2004:

* Delaware 31% to 27%
* Indiana 42% to 36%
* Maryland 28% to 23%
* Nebraska 41% to 36%
* North Carolina 40% to 37%
* Oklahoma 43% to 39%
* Pennsylvania 30% to 27%
* South Dakota 39% to 35%
* Virginia 38% to 33%

Actually 9 is not a bad number, even if many were irrelevant downturns in heavy red states or solid blue states. Note the 5% drop in Virginia. That was most significant of all because it verified the legitimate ideological shift that was being heavily touted at the time.

This time we may get that type of Virginia shift in states like Texas and Arizona. I would be more surprised if it happened in states without increasing Hispanic presence. Overall, as I emphasized in the other post, I'd like to prioritize Florida and not have to rely on flails in heavy red territory. If Biden's national edge drops to let's say 4 points then states like Texas and Georgia are long gone but we still could pull it out via Florida and the midwest.

BTW, this stuff became more uncertain after 2008 because not every state is exit polled. From 1992-2008 every state and the District of Columbia were sampled. Immensely helpful. But beginning in 2012 roughly 20 states were dropped in a cost cutting measure.