A few thoughts on the state of the presidential race
By 
There’s a new USA TODAY/Suffolk University poll (highly rated in the 538 curation), which shows an eight or nine-point swing in favor of Harris, relative to the last survey it took at the end of June, right after the debate that, at the moment at least, looks like it may have cost Trump the election:
Democrat Kamala Harris has surged ahead of Republican Donald Trump, 48%-43%, a new USA TODAY/Suffolk University Poll found.
The findings reflect an eight-point turnaround in the presidential race from late June, when Trump had led President Joe Biden in the survey by nearly four points.
The vice president’s small lead was fueled by big shifts among some key demographic groups traditionally crucial for Democrats, including Hispanic and Black voters and young people. Among those with annual incomes of less than $20,000, in the biggest change, a three-point Trump edge over Biden in June has become a 23-point Harris advantage over Trump in August.
The poll of 1,000 likely voters, taken by landline and cell phone Sunday through Wednesday, has a margin of error of plus or minus 3.1 percentage points. With the election approaching, the survey is now measuring likely voters; previous polls were of registered voters.
When Biden was doing badly in the polls, there was a lot of copium on LGM about how the polls are meaningless/slanted intentionally toward Republicans etc. Polling is a very inexact science of course, and some polls are intentionally dishonest, but most aren’t, and honest pollsters are constantly trying to refine their models to make them more accurate given constantly changing conditions.
The 2020 polls ended up overstating Biden’s support by about three and half percentage points at the national level. The 538 average is pretty sobering in this regard, as a final 8.4% margin (it had been 10.5% in mid-October when Trump’s campaign seemed to be melting down over the Access Hollywood tape eta: oopsie that was 2016; how time flies!) ended up being 4.5% on election day. A similar though not quite as stark dynamic played out in 2016: for whatever combination of reasons, Trump’s support in the polls in both years turned out to be underestimated. I suspect this is because Trump brings out a lot of people who don’t traditionally vote, and are therefore hard to capture in the polling models.
Some general thoughts:
(1) It’s just stunning to me that Trump is going to get 45% of the vote. That’s his realistic floor. That’s with maybe 2% going to third parties, so that would be an eight-point popular vote win for Harris. Again, that’s best case scenario.
(2) It’s going to be close in the swing states. I mean if Harris wins by eight percentage points in the popular vote she’s going to win, but if she wins by five, which means around eight million total votes, she might not. I’m so old I remember when plenty of Reasonable Moderate Democrats used to defend the Electoral College, which right now is basically a national election cheat code for rural white religious nut cases.
(3) I think it’s all going to come down to turnout and enthusiasm. There are very few real swing voters any more, because the kind of person who has trouble choosing between the Democrats and the Republicans is somebody who probably can’t tell you what country is directly north of the USA.
(4) The enthusiasm for Harris is driven by hope; that for Trump is driven by fear.
I’m reminded of the line from A BronxTale where the kid asks the mob boss whether he would rather be loved or feared, and he says ideally both, but if he has to pick one he would pick fear. The kid asks him why, and he says “Fear lasts longer.”
To end on an optimistic note, somebody pointed out to me that the Democrats’ GOTV efforts in 2020 were badly hampered by Covid, and that the few Democrats who did large amounts of in person GOTV campaigning did better than their polling averages, while those that didn’t did worse. That won’t be a factor this time, and I do suspect that it will all come down to GOTV efforts — that is, to maintaining enthusiasm and energy, as opposed to changing the minds of undecideds/soft Trump voters.