Seems to me there’s a much greater chance than usual of the polls being way off this time. That Nate Cohn article from a couple weeks ago talked about how pollsters are weighing samples toward the last election (essentially causing a big herding effect), even though that’s been bad practice every previous election.
The pollsters who aren’t doing that are sometimes showing weird changes, and there’s some evidence that we could see big movement among certain groups (minorities toward Trump, whites toward Harris). If any of those trends are real, pollsters are going to have a bad time.
Add in non-response bias getting worse and worse and yeah, I wouldn’t be surprised if the polls end up being off by like 5% this time. In which direction, I dunno.
Have a very quick and shitty model of the 10 largest counties in Georgia PLUS 5 significant rural >+60 Trump counties. Looking at disparity between total 2020 votes cast vs votes recorded to date this year to estimate votes for each candidate this year (assuming 2020 margins hold across the counties). Basically, trying to isolate turnout to date by county.
Under these assumptions, I see a 0.33 percentage point shift towards Trump.
How are things trending? Well, another metric is votes cast 10/15 vs 10/21.
For the 10 largest counties in Georgia (which went for Biden 63.33% in 2020), 119,062 votes were cast 10/21 compared with 154,460 cast 10/15, a decline of 22.9%. For all other counties (which went about 65% for Trump in 2020), 128,743 votes were cast 10/21 compared with 158,939 cast 10/15, a decline of 19.0%. Favors Trump once again.
Random thought: its always kind of dismal to do an analysis and at the end go, “Hmm, I really hope this whole analysis was meaningless and it’s not suggesting what I think it’s suggesting.” Because that’s where I’m at with this.
One of the benefits of the massive polling fuck ups of the past two cycles is that we are starting to get somewhat frank depictions of how the sausage gets made with these things and that polling is as much punditry as anything else. With response rates approaching zero and results prone to massive systematic bias, pollsters basically decide in advance who they think will vote and how and then go out and look for confirmation. And failing that they just weight the results to match their preconceived notions about the electorate.
To a certain degree it is not clear to me the benefit of actually doing the poll itself. I am told that the big problem in 2016 was that the models were not weighted by education. Most pollsters didn’t do that because they didn’t think that party preference would correspond to education. But they had the data right there! Isn’t the point of polling identifying exactly that type of pattern! If you are going to collect that stuff and then ignore it because you didn’t think it was that important at the outset, why are you bothering all those people at home? Some of them are trying to eat! Just tell us who you think will win and move on with your life imo.
Speaking of Nate Cohn, another good article by him today discussing two theories as to why polls have underestimated Trump during election years (but not during midterms).
The first theory, which seems quite plausible to me, is essentially that low-engagement, low-info people tend not to respond to pollsters. And they tend to vote for Trump.
This is another reason why I hate when democrats paint Trump as an existential threat to democracy. Telling everyone that this is The Most Consequential Election in History (while possibly true) is bad for democrats! We need Joe Rogan bros to be demoralized and to stay home. Getting them riled up means they vote Trump.
Calling Trump a weirdo and making fun of him was working. Calling him the big bad wolf is not.
Actually, if we believe anything he says (I dont) lately he has been signaling that a right wing chud wont end up purchasing all his assets and giving them back to him in 4 weeks.
Also, every time he tries to make a new online store using his name or products the judge tells him to knock it the fuck off. Im just waiting for him to try to do it some point and stepping over the line into criminal liability and not just civil.
He has the brand until November 20th (or somewhere around there) where it is being forced into sale. Word on the street is there is big money lining up on both sides to buy it.
At this point I’m hoping that the vibes seeming off is just a function of most people who are voting D being completely checked out.
I haven’t really been paying much attention this whole cycle. I’m not going to learn anything new that would sway my opinion, and it’s actively bad for my mental health to consider another Trump Presidency. So all I can do is hope for the best and focus on things in my life that I can control or that give me joy.
I thought the same thing, hopefully its along the lines of the early voters have made their decision and dont want to think another second about this shit until its over and the polls arent reflecting that
Agreed. But if you stop saying “existential threat to democracy” you have to give people something to vote for instead, and dems seem very reluctant to do that.
I think the goal was to drift rightward and try to run out the clock–election version of a prevent D. Not sure that’s the best approach against Trump.
I dropped my ballot off this morning at an empty library parking lot. Someone else dropped off two ballots just before me. Ballot stuffing witnessed! Vote by mail really is great though, joking aside!
Voted yesterday in a small bright red FL county. There were maybe 5 people ahead of me in line to get their ballots and another dozen or so in the booths. Left work at 4 and walked out of the voting site at 4:15. Got to the Publix deli by 4:22, there were 5 people in line for subs so I grabbed wings instead.
Please read into to my anectode as best fits your needs.
