PA is a big part of the more likely ones, though.
1) Is 1000 likely voters enough?
2) It's funny how there's no mention of The Economist's track record on such things.
3) Along the same lines there's no mention of magin for error / level in confidence in this *prediction*. Oops I mean projection.
4) While there are plenty who are "out" (so to speak about supporting DJT), there are still those who are not so open. How do these polls account for these people?
2) The Economist's prior election models are still on their site, so you can judge for yourself. They don't have a huge history - 2020 was the first year they did a full model, I believe.
3) The level of confidence is basically the 3 in 5, etc. https://www.economist.com/interactive/us-2024-election/predi...
4) Are you talking about the poll or the economist model? You're going back and forth here. If it's the economist's model, polls are only one input and they try to balance it against other historically accurate indicators, see above. If you mean the poll, it depends entirely on the specific pollster
The article is also out of date on Silver's model, fwiw. He now has Harris at a 58% chance to win the EV, after having had Trump as the winner there for basically the entire race.