Morning update: Texas reporting 9,669,246 votes cast at the close of early voting. That's 57% of registered voters and 700,020 more votes than were cast in Texas in *all of 2016.*
Largest increases in 2020 early votes cast vs. 2016 total votes cast (TX counties w/ 100k+ registrants only):
The final day of TX early voting was a relatively strong one in the most heavily Hispanic counties. Total votes cast in the six most populous border counties (El Paso, Hidalgo, Cameron, Webb, Starr, Maverick) jumped 9.3%, vs. 6.2% everywhere else in TX.
It's quite possible Texas will cast more than 12 million votes in this election. That would be a massive increase after 8.96 million votes cast in 2016 and 8.37 million cast in 2018.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Morning update: Texas just surpassed 95% of its 2016 total votes cast w/ two days of early voting & Election Day left to go. We're just headed for a massive, unprecedented turnout there (and a lot of other places).
Top TX counties in current '20 votes cast vs. '16 total votes cast (pop. growth a factor here):
In Sumter Co., FL, as of this morning 72,695 out of 105,612 registered voters (69%) have already voted - which I believe is the highest turnout rate outside of auto-VBM states.
Although we know that Trump is likely to better in the EDay vote than the early vote virtually everywhere, there's not *that* huge a difference between the Sumter voters who have voted so far and those who haven't...
Among the 72,695 Sumter voters who have cast ballots, they break down:
58% R
25% D
17% NPA/Other
Here's the breakdown of the 32,917 Sumter registrants who *haven't* yet cast ballots:
Tfw you've been devouring hundreds of district-level polls for decades and everyone has suddenly woken up to how instructive they can be.
The truth is, the majority of presidential race polling is conducted at the *district,* not state or national level.
Most of it is never made public by parties/PACs, but is of high quality b/c it's conducted to make critical resource allocation decisions in House races.
Four years ago, these district-level polls showed flashing red warning signs for Clinton in heavily white working-class districts (see below), but many people overlooked them.
Breathtaking: statewide, Texas just surpassed 80% of its total 2016 votes cast, leading the nation. And there's still more than a week of voting to go.
Also, tbh the dropbox issue was never going to have that big an impact on turnout b/c a relatively tiny share of TX voters is eligible to vote by mail to begin with.
You know what will have a big impact on TX's outcome? Turnout in the heavily Hispanic border counties, which is currently at ~32% of registered voters compared to 50%+ in TX's major metro suburbs.
On one hand, TX isn't (and was never) poised to be a "tipping point" for 270 EVs. On the other, even a modest investment in several smaller, low-turnout TX media markets could now determine which way TX tips.
Here are my updated demographic splits in average crosstabs of live-interview national polls, w/ new @QuinnipiacPoll thrown into the October mix.
A few breathtaking splits:
- Biden now +25 w/ women
- Biden now +10 w/ seniors
In my view, the biggest "tell" Trafalgar etc. are off the mark? Private, district-level GOP polling conducted to make spending decisions backs up what we're seeing in averages of live-interview polls: Trump doing terribly w/ college whites/women into October.
The only group where we're seeing improvement in Trump's margin vs. '16? Hispanic voters.
Today, live-interview average has Trump down 27 points, vs. 38 points per @UpshotNYT's average of final '16 polls.