Real Close

9:36 PM: I mentioned earlier that we've clearly had a pretty substantial polling failure. One thought occurs to me. One of the most oft-repeated stories in the history of polling was one of FDR's races - I think 1936. The polls of the day completely got it wrong. If I recall the reason was that the pollsters didn't figure in that a lot of poorer voters didn't have phones. So phone polls had a very Republican skewed sample of voters. Now, that seems obvious. But it hadn't shown up in the few earlier elections that had been polled because they hadn't been as clearly polarized by class. In other words, you had a way of predicting elections that seemed to work. But it was based on a certain kind of political configuration, a certain social basis for divisions between the parties. That system worked in the pre-New Deal politics. But it failed miserably in the different politics of the New Deal. And it only became clear that the old model was based on a certain kind of politics when it failed. We need some explanation for why this result is so different from what was expected.

9:30 PM: In case you're having a difficulty putting all these moving pieces together, this is now extremely close nationwide. Trump is running strong in many blue states. He could definitely win the presidency. I'm not saying that's going to happen. But he's competing in a lot of places that Clinton needs to win.

9:24 PM: We've had many conversations about polling failures. It seems like we've definitely had one here. We had lots of reporting about a big Latino vote surge. It's still not clear what happened there. I'm not sure that reporting was incorrect. It may simply have been matched and overcome by a surge of conservative white voters. But a lot of polls and a lot 'models' were significantly off. There's no getting around that.

9:14 PM: I haven't been doing much blogging here because I'm mainly looking at the map and trying to see what's happening. It is definitely not what we expected. Trump is outperforming recent Republican candidates in rural America. And it's putting him ahead or close in a lot of states. For the moment he's ahead in Virginia. You always have Northern Virginia reporting late. But Virginia shouldn't be with Trump in a lead with this much of the vote reporting.

9:05 PM: There are a lot of states that we thought Clinton would take early that Trump is running really close.

8:53 PM: We don't know how this will play out. But the big story right now in Florida is that Trump is really outperforming Romney in the Republican counties.

8:51 PM: This could go really long in Florida.


Josh Marshall is editor and publisher of