What’s the point: Let’s state it up front: winning the popular vote means nothing in and of itself. You only become president if you win the Electoral College.
That said, the national polling matters a lot. We have a ton of it, unlike in a number of swing states. It’s more accurate than the state polling on average. Most importantly, we have a fairly good idea of what Biden’s margin needs to be nationally in order to win in the Electoral College.
There is a limit, however, to the extent to which Biden can win in the popular vote without also taking an Electoral College majority. The nation, after all, is made up of the states that also make up the Electoral College. If the margins move in the swing states, then they’re very likely to move nationally as well.
If Biden wins by more than 5 points nationally, he’s almost certainly going to win enough electoral votes to get to at least 270 and take the Electoral College. There’s simply little record of a difference between the margin in the key swing states and the national vote being greater than 5 points.
If Biden wins by somewhere between 3 and 5 points nationally, he’ll be the clear favorite in the Electoral College, even if there is some non-negligible chance Trump could emerge victorious.
Biden’s national advantage right now is clearly outside this 5-point window.
There would need to be a polling miss for Trump to have a realistic shot in the Electoral College. A 2- or 3-point miss in the national polls would be far from unheard of, but it’s not as likely as you might think, at least in comparison to state polling.
For a number of reasons, including that we have a lot more data to work with, national polls have been more accurate than state polling since 1972. We obviously all remember the state polling errors in 2016, but it goes back further.
During the same period, there has never been a year in which a simple average of the national polls has been off by more than 6 points. (You’d have to go back about 70 years for that.) The 95% confidence interval for state polling, meanwhile, has been closer to +/- 9.2 points since 1972.
The point is that a national polling miss isn’t as likely as a miss in key swing states. If Biden is up by the same margin nationally on Election Day as he is now, he’ll be a heavy favorite to win not just nationally, but in the Electoral College as well.
But it’s more than just the polls on Election Day for why we need to pay attention to the national polls. It’s about the period before it.
Nationally, we’ve had nearly 20 different pollsters release results this month already. Some of them have put out multiple polls already. And because we’ve had so many polls nationally, we have a very good idea of where the race is. Biden’s ahead by somewhere around 7 to 8 points, depending on how you average.
There will be a number of national polls released every week. This will allow us to know if the race is really shifting in a way the state polling likely cannot keep up with.
For now, the race remains Biden’s to lose.