Joe Biden is leading in the national polls, big league. He’s leading in most of the swing-state polls. States that were safe for President Trump are now in play.
On the other hand, Trump outperformed the polls four years ago, and he could gain on Biden for one reason or another over the next 3 1/2 weeks. And, of course, Trump doesn’t have to win the national popular vote (there’s no such thing!); he just has to win states worth 270 Electoral College votes.
So, how much ground does Trump need to make up? It’s not a simple question, but here’s one (admittedly simplified) way to look at it.
If FiveThirtyEight is correct in its relative ranking of the states, either Pennsylvania or Florida will be the “tipping point” state.
That is, for Trump to win, he needs to hang onto Iowa and Georgia and also overcome his deficits in Ohio, North Carolina, Arizona, Florida, and Pennsylvania. It’s possible that he could win without Florida or without Pennsylvania, but that would require him winning Wisconsin plus Nevada, plus either Nebraska’s or Maine’s 2nd Congressional District.
- In Pennsylvania, Trump is down 7.1 percentage points in the RealClearPolitics average.
- In Florida, he’s down 4.5 points.
- In Wisconsin, he’s down 5.5 points.
- And in Nevada, he’s down 6 points.
Trump’s deficit is smaller in Ohio, North Carolina, and Arizona.
Trump is unlikely to surge in any one of those states without surging in all of them, and so the easiest path to victory, if we use the RCP average, is to win Ohio, North Carolina, Arizona, Florida, Wisconsin, and Nevada plus Maine’s 2nd District. This would give Trump exactly 270 electoral votes, the bare minimum.
The toughest climb in that route is the 6 points in Nevada. If Trump gains 6 points there, he’s likely to also gain about as many points in the other five states where he needs to make a comeback. So, the simplest answer to “how far behind is Trump really?” is 6 points, because that’s his deficit in Nevada.
But FiveThirtyEight believes Trump actually has a better chance to win Pennsylvania than Nevada. The forecast vote share shows Trump down 6.1 points in the Keystone State. So, on either of these two paths, it looks like Trump would need to gain about 6 points.
You’re probably thinking, but weren’t the polls wrong four years ago? Yeah, kind of. But that’s not the great news for Trump you might think it is. First off, in Nevada, which is famously hard to poll well, Trump actually underperformed his 2016 polls: He was beating Hillary in the RCP average but lost the state.
Secondly, many pollsters have learned from their 2016 mistakes — most importantly, they now make sure non-college-educated people are properly weighted in the polls.
In any event, Trump overperformed the RCP average in 2016 by 2.8 points in Pennsylvania, 1 point in Florida, and a stunning 7.2 points in Wisconsin.
If you assumed the polls would be just as off this year as last time, and so gave Trump those 2.8 points in Pennsylvania, he has 3.3 points to gain there. Again, he needs to get about 6 points in Pennsylvania to win.
Now, there are a thousand caveats to give to this approach, the biggest being that these averages don’t fully reflect Trump’s coronavirus diagnosis and subsequent behavior.
