With polling of the 2020 race just about concluded, former Vice President Joe Biden is in an interesting position.
Biden’s national lead is large, and he has consistently led polls in states that would be sufficient to deliver him 270 electoral votes and therefore the presidency — most notably, the triad of Michigan, Wisconsin, and Pennsylvania.
But Biden’s lead in Pennsylvania — 1.2 to 4.7 percentage points, depending on the polling average — is not quite big enough for Democrats to be completely confident in it, particularly given what happened in 2016, when Donald Trump won the state and the polls underestimated Trump’s performance there by about 4 points.
Biden is also either narrowly ahead or about tied in polls of another set of swing states that he doesn’t even need to win: Florida, North Carolina, Arizona, Georgia, Iowa, Ohio, and Texas. His leads are small (1 to 3 points in polling averages), and the others are pure toss-ups. If Trump sweeps these states or comes close to doing so, Biden really does need to hang on to Pennsylvania.
Still, it’s hard to point to many bad signs in the final polls for Biden. He’s clearly the favorite, with a 9 in 10 chance of winning, according to FiveThirtyEight. Indeed, if the polls are underestimating Biden’s strength by just 2 points, he’d win all the swing states listed above, and finish up with a 400+ electoral vote landslide.
And yet due to the Electoral College, Biden’s leads are not enough to dismiss Trump’s path to victory entirely. To get 270 votes, Trump needs to come out on top in nearly all the toss-up states, and snag Pennsylvania as well.
Trump probably needs to win Pennsylvania to have a shot at victory
To get a sense of where the polls are, let’s start off with a scenario where Biden and Trump each win everywhere they’re up by 4 percentage points or more, according to the FiveThirtyEight polling averages.
In this scenario, Biden would win the states Hillary Clinton won in 2016, plus Michigan, Wisconsin, and Pennsylvania. That would be enough for him to win the presidency without needing any of the toss-up states that polling averages show being closer (depicted in gray on this map).
So Trump very much needs to find weak spots in this map. And, understandably, he’s focused on the same weak spots that were in Democrats’ “blue wall” in 2016 — Michigan, Wisconsin, and Pennsylvania.
The final polls showed Hillary Clinton winning all three of those states, but Trump won them all instead, each by a margin less than 1 percentage point. However, Biden’s poll leads in each state are bigger than Clinton’s leads were in 2016.
In, Michigan and Wisconsin, Biden leads by about 8 percentage points in the FiveThirtyEight averages. The RealClearPolitics averages show it a bit closer, with Biden’s lead at 5 to 6 points in each. But basically, Trump has to hope for a fairly large polling miss to put him in contention in either state.
Pennsylvania, however, is a bit of a different story. FiveThirtyEight places Biden’s lead there at 4.7 percentage points, while it has dropped in the RealClearPolitics average to 1.2 percentage points (including some less well-known and partisan pollsters). The point is, Pennsylvania appears to be closer than Michigan or Wisconsin. It’s also the biggest of the three states, with 20 electoral votes at stake.
So if something were to go wrong for Joe Biden, it would probably entail a loss in Pennsylvania. Conversely, if Biden wins Pennsylvania, he’s likely won the presidency as well.
Trump would also need to win nearly all the toss-up states
Let’s return to this map, of states where each candidate is leading by 4 points or more in FiveThirtyEight’s averages, and focus on another takeaway: Trump is only up to 125 electoral votes in it, less than half the number he needs for victory.
So to get anywhere even close to winning, Trump needs to win the vast majority of votes in the toss-up states — those where neither candidate is ahead by four points or more. And the bigger the state is here, the more important it is for Trump’s math.
Texas, with 38 electoral votes at stake, is clearly the most important. But FiveThirtyEight shows Trump ahead by a mere 1.1 percentage points in the traditionally Republican state, and RealClearPolitics shows Trump up by 1.2. Trump absolutely cannot afford to lose this one.
Florida, with 29 electoral votes, is also essentially a must-win state for Trump. Here, FiveThirtyEight shows Biden with a 2.5 percentage point lead, while RealClearPolitics shows it closer, with Biden up by .9 points.
But there’s more.
- Ohio has 18 electoral votes, and it’s quite close (Trump up by 0.8 per FiveThirtyEight, Trump up by 1.4 per RealClearPolitics).
- The same is true for Georgia and its 16 electoral votes (Biden up by 1.2 per FiveThirtyEight, Trump up by 0.2 per RCP).
- And there’s also North Carolina with 15 electoral votes (Biden up 1.8 per FiveThirtyEight, Trump up by 1 per RCP).
- Iowa’s 6 electoral votes are only likely to be decisive in a very close contest, but it certainly doesn’t help Trump if he loses them, and he’s only up by about 1 or 2 points in both averages.
So let’s say Trump pulls it out in all these states — Texas, Florida, Ohio, Georgia, North Carolina, and Iowa. Let’s give him Pennsylvania and the lone electoral vote in Maine’s second district, too. Then the election comes down to the one state remaining: Arizona and its 11 electoral votes.
Arizona is another traditionally Republican state, but it’s been one of the strongest of this group of swing states for Biden (though it’s still quite close) — Biden leads by 2.6 percentage points there per FiveThirtyEight, and by 0.9 per RealClearPolitics.
So, interestingly, Arizona could be a “Plan B” for Biden if he loses Pennsylvania. Winning it would give Biden 270 electoral votes, exactly what he needs to win — though he’d have to hold on to Nebraska’s second district (which Trump won last time but where polls show Biden leading this time), and to prevent any defections from faithless electors. (If neither candidate gets to 270 electoral votes, the election will be determined by votes of state delegations in the new House of Representatives in January — and it’s unclear which party will control more of those.)
Meanwhile, if Trump flips Arizona in addition to Pennsylvania, here’s his victory map:
You can see Trump has little room for error. Of the toss-up states, he can afford to lose Iowa, but if he loses any others, he’ll have to make up for those losses by plucking away more states where Biden leads by a lot (like Wisconsin or Michigan).
How different is this from 2016’s final polls?
This roundup may be giving you a sense of déjà vu — since the polls just before the 2016 election also found that the Democratic candidate looked to be ahead in enough states that would deliver victory.
There has been much digital ink spilled about how 2020 is not 2016, and there are indeed many differences. This time around, Biden is leading by more nationally than Clinton was. Polls also show Biden leading in more swing states, usually by bigger margins, compared to Clinton. Biden tends to top 50 percent in more state polls as well, since there are fewer undecided and third-party voters. Analysts with access to non-public polling of congressional districts report it generally looks good for Biden.
Yet there is one similarity: In the likely tipping point state, Pennsylvania, FiveThirtyEight’s average puts Biden ahead by 4.7 percentage points — and it had shown Hillary Clinton ahead by 3.7 percentage points. (Trump won by 0.7.)
There is no reason to necessarily expect the mistakes of 2016 to be repeated. Polling error could also underestimate Biden’s strength. And remember that Pennsylvania alone wouldn’t be enough — Trump probably needs to win all of Florida, North Carolina, Georgia, Ohio, and Texas as well. Overall, though, this is why Biden looks to be in a strong position — but there is still just a hint of doubt about what will transpire.