Tuesday, October 27, 2020

One week to go

 Some punditry thoughts:


Looking at the last 4 weeks of this weighted polling average from 538, the simplest explanation of the movement is that 1%-2% in the undecided middle shifted to Biden and stayed with him, while 1%  drifted from Trump to undecided and then drifted back. The changes started around Trump's disastrous first debate and his getting covid. They partially match a pattern over the last four years where Trump does something that even some of his base dislikes and becomes even more unpopular than usual, then gradually drifts back to baseline, fairly unpopular level.

This doesn't totally match that usual pattern, in that Biden hasn't lost the voters he gained. It could be due to undecided voters feeling like they need to make up their minds and stick with it, or even having set their opinion in stone for a while by having already voted. One other possibility is the third covid wave, which started nationally about three weeks ago, and earlier in a few states. There's not a lot of obvious evidence, yet, of it affecting the polls, but maybe Biden wouldn't have been able to keep his new supporters absent the third wave reinforcing his message that Trump mishandled the pandemic.

In this last week before Nov. 3, Trump is way down in the polls, he's got much less money to advertise than Biden, and covid is really getting going. I think hospitalizations in particular are media- and especially television-friendly, and the rise for them isn't yet at record levels but is enough to take media attention. Some covid surprise from Pence or other prominent person on the Democratic side could shake up things.

The Electoral College is undemocratically shifted about 3-4% towards Republicans, and a near-extreme-case-scenario is polls underestimate Trump by 4%, so if polls drift down to an 8% differential then Trump has an outside chance. Except that a lot of people, maybe half, will have voted by Election Day. I'd guess Trump needs polls to move to a differential of 7% to 7.5% by Election Day to have a chance of winning the Electoral College.

This is assuming attempts to stop the vote count after Election Day are unsuccessful. Hard to predict what would happen in that case, to say the least.