According to , Donald's support with Republicans went up to 72% (+5%) after Sunday's tweets. Support by independents was down 10% and support among Democrats dropped 2%. Eric Trump claimed his support among Republicans was 95%, but, you know, Eric.
Ordinarily, I'd say Democrats are going to be Democrats and Republicans are going to be Republicans and look mainly at the Independent numbers. But the tweets were so wrong that I'm led to ask "What in the hell is wrong with Republicans?" But then I remember that the GOP is down to only 26% of registered voters, these days, as centrists have left the party. So, I'm thinking that a far more interesting survey would be one that breaks people down by what party they were a member of 5 or 10 years ago. Moderates have left both parties but I can't think of anything that has driven moderate Dems away like Trumpism has with Republicans.
Then again, maybe more extremists have been joining the GOP. In an ideal world, Independent-only numbers should be about halfway between the two, but they never are. They always skew toward the Democratic numbers - or maybe a better way of putting it is that they skew away from the Republican numbers, because the Republican numbers are usually the complete opposite of the Independent numbers - even on things that are accepted as facts. I don't have any numbers handy, but I know that on the question "Did Russia try to meddle in our election," the Republican responses always skew heavily toward "No."
Long-winded way of saying that on opinions, at least, reading the partisan numbers will only make you crazy. "Americans in general" seems to be more meaningful than even the Independent numbers, these days. But, using the "Five years ago" metric would be a lot more interesting. Almost as interesting as a nationwide survey that shows opinions by congressional district, although that would be impossible. But it sure would be helpful in trying to understand some of the opinions.
Don't look at me, I just work here.