The most puzzling thing for me in this election (and the reason the scenario of Rs barely keeping/just barely losing the House keeps nagging at me) is this: the overall atmosphere isn't that atrocious for Republicans, and a lot of "big picture" signifiers of waves aren't here. 1/
Like, if you compare it to 2006, Trump's job approval is substantially higher than Bush's 37%. The economy is significantly better. Republicans have about the same amount of baseline exposure in the House (i.e., Rs in D-leaning seats). 2/
We haven't seen a massive surge in Democratic party identification. The enthusiasm gap is present, but not really of the magnitude that we've seen in the past. And we typically see a big break at the end. Maybe by the end of the day I'll eat my words here 3/
But while I think this week's polls have been worse for Rs than last week's, we don't seem to have a massive, late shift. We have more than enough seats in play for Ds to take the House, and ultimately I think there's just too many fires for Rs to put them all out. 4/
But we also have a LOT of House seats that are polling within a point or two. If we get even a *slight* break among undecideds toward Republicans they end up keeping a lot of those seats. Of course, the opposite is true as well. 5/
But a *lot* of those seats also have some serious Republican DNA in them so I'd guess an R break is more likely than a D break(?) Anyway, this is what keeps me awake at night. If Rs end up holding, or losing by a seat or two, I think I'll look at this thread 6/
and think "what the hell was I thinking?" Don't even get me started on the Senate, where I sort of feel like it will either be R+0 or R+5 without much room in the middle. But I guess that is what makes this fun. 7/7
Oh FFS people. Silver's polls-only model gives Rs close to a 1-in-4 chance of holding the House, and his overall model gives Rs roughly the chance of having three kids and having them all boys. Which I promise you happens. 8/
Yet the mainstream Twitter hivemind seems almost unanimous in its assessment that Ds will take the House. If there's any lesson to be learned from 2016, it's that this should make us nervous. 9/
Like, the way to do this stuff is NOT to sit around and look for reasons you could be right. It's to look for reasons that you could be wrong. That's part of the exercise I put myself through constructing scenarios with Ds winning 40 and Ds winning 20. 10/
And it is surprisingly easy to construct either one. My hard count is D+32 right now, and I think that's about right but man. There are plenty of reasons to put it on the high side, which right-thinking Twitter has been dwelling on ad nauseum, 11/
But there are also lots of indicators as to why it might come out on the low side as well. Again, if you learned nothing else from 2016, it should be that we ought to pay attention to those as well.
