So let's talk numbers shall we? There's a lot of fretting and nailbiting about the potential for Trump to win a second term. I've been guilty of it myself, so I decided to run the numbers and see just how much ammo I need to stock and/or which border crossings are the closest. Turns out...it's not that bad.
It's no secret that many (the majority, even) of state elections for President are non-competitive. New York hasn't voted for a Republican since 1984, when Reagan annhilated Mondale. Oklahoma hasn't voted for a Democrat since 1964. So most Presidential elections come down to the handful of states that are in fact competitive. This is why it seems like the fate of the country always winds up depending on Florida or Ohio. The question then becomes, how many "in-play" states are there, and how many does Biden need to win?
The answer is that there are actually more than usual this cycle, and not that many. For all of Trump's braggadocio about his victory, it was pretty damn weak in the grand scheme of things. Out of 58 elections, Trump's ranks 46th in terms of Electoral College margin of victory. 304-227 is a 77 point margin, meaning that Biden only needs to flip 39 electoral votes (and defend Clinton's wins from 2016) to win. Almost all the competitive states this cycle were red states in 2016 and have double-digit electoral votes. So he only has to flip 2-3 of them to squeak out a win.
Let's start with the non-competitive states and see where each side starts at. To determine non-competitive, I'm looking at aggregate polling totals, as determined from RealClearPolitics and 538. Anything where one candidate leads the other by 7 or more points is deemed non-competitive. In cases where no polling has been done, past history is enough to determine the likely winner (aforementioned examples of New York and Oklahoma, for examples).
Safe Trump Wins (and their electoral vote value, and if available the current polling margin):
Alabama (+18.6, 9 EV)
Idaho (4 EV)
Indiana (+14, 11 EV)
Kansas (+7, 6 EV)
Kentucky (+9, 8 EV)
Louisiana (+16, 8 EV)
Mississippi (+15, 6 EV)
Missouri (+7.7, 10 EV)
Montana (+11, 3 EV)
Nebraska (5 EV*)
North Dakota (3 EV)
Oklahoma (7 EV)
South Dakota (3 EV)
Tennessee (11 EV)
Utah (6 EV)
West Virginia (5 EV)
Wyoming (3 EV)
TOTAL: 17 states, 108 electoral votes
Safe Biden Wins:
California (+30.3, 55 EV)
Colorado (+10, 9 EV)
Connecticut (+23, 7 EV)
Delaware (+21, 3 EV)
District of Columbia (3 EV)
Hawaii (4 EV)
Illinois (20 EV)
Maine (+11.5, 4 EV*)
Maryland (10 EV)
Massachusetts (+35.7, 11 EV)
New Hampshire (+9.7, 4 EV)
New Jersey (+18.8, 14 EV)
New Mexico (+14.5, 5 EV)
New York (+26.3, 29 EV)
Oregon (+12, 7 EV)
Rhode Island (4 EV)
Vermont (3 EV)
Virginia (D+12.5, 13 EV)
Washington (+34, 12 EV)
TOTAL: 19 states, 217 electoral votes
*Nebraska and Maine both have a proportional allocation system where a single electoral vote is awarded to the winner in each congressional district and the remaining votes are awarded to the statewide popular vote winner, but ultimately it's not worth the effort to fiddle about determining -- Biden might pick up a vote in Nebraska, Trump might pick up a vote in Maine. Ultimately, it's highly unlikely to come down to a 1-2 vote margin.
So we see here two big advantages the Democrats have:
1) Big safe states. It only took the Dems 4 states to surpass the "safe" Trump votes: California, Illinois, New York and New Jersey (118 EVs). And as you can see from the polling numbers and historical voting record, those are very safe states for them. MOST of their safe states are very safe (20+ point margin). Whereas even the deepest red states don't crack 20+.
And before someone complains about the tyranny of the popular vote, please note that those four states together are roughly 80.4 million people, whereas the entirely of Trump's 17 "safe" states total 55.6 million. Republican states typically always have a lower population-to-EV ratio because 3 is the minimum number. If the 538 EVs were strictly proportional (~600k per EV), Wyoming would get less than 1 vote.
2) Better than a 2-to-1 advantage off the starting blocks. They only have to pick up 53 votes, Trump has to get 162.
There's also a third advantage, which is that they're "defending" very little territory. Let's look at the "in-play" states (a number of which I will shortly remove and hand to Trump for reasons which we'll discuss)
These I'll list with the current polling margin, EVs and who won it in 2016.
Alaska (R+3, 3 EV, Trump)
Arizona (D+5.7, 11 EV, Trump)
Arkansas (R+2, 6 EV, Trump)
Florida (D+1.2, 29 EV, Trump)
Georgia (R+1.3, 16 EV, Trump
Iowa (R+1.7, 6 EV, Trump)
Michigan (D+4.2, 16 EV, Trump)
Minnesota (D+5, 10 EV, Clinton)
Nevada (D+4, 6 EV, Clinton)
North Carolina (D+0.8, 15 EV, Trump)
Ohio (D+2.4, 18 EV, Trump)
Pennsylvania (D+4.3, 20 EV, Trump)
South Carolina (R+5, 9 EV, Trump)
Texas (R+3.5, 38 EV, Trump)
Wisconsin (D+6.4, 10 EV, Trump)
TOTAL: 15 states, 213 EVs
Observations:
1) That's a lot of votes still "in-play", but they're almost entirely votes Trump needs to retain. The only states Dems are defending from 2016 are Minnesota and Nevada, and they're doing well there. The most recent polls have Biden up 9 in Minnesota (and yet RCP downgraded it from "Leans Biden" to Toss-up, because...
) and up 4 in Nevada.
2) The fact that states like Alaska, Texas and South Carolina are even on this list should scare the beejezus out of Trump's campaign.
South Carolina has not voted for a Democrat since 1976, when Jimmy Carter was able to capitalize on his Southerner status. Carter remains the only Democrat to win South Carolina since the Great Realignment of 1964/68. The closest any other candidate has come was Obama in 2008, with 44.9% of the vote. Biden's pulling 46% right now, +/- 4%. Part of this may have to do with the Dems running a fairly popular African-American candidate (Jaime Harrison) against Lindsey Graham, who has managed to piss off both the Trump crowd and the Democrats by opposing Trump enough to irk the MAGAts, but caving when it's important. I still don't think Biden will win South Carolina, but it's breathtaking that Trump even has to defend it.
Alaska goes one better than South Carolina, and hasn't voted for a Democrat since 1964. Alaska told Jimmy Carter what he could do with his nuts and only gave him 35.6% of the vote. Likewise, during Obama's historic 2008 wave, he only got 37.89% of the Alaska vote, no doubt due to the presence of Alaska's own Caribou Barbie on the Republican ticket. The fact that Biden is within 3 of Trump here is remarkable. But again, I expect Alaska to go to Trump.
Texas -- everyone keeps hoping for a tectonic shift in the Lone Star state, driven by changing demographics. Trump only got 52.1% of the vote in 2016, the lowest percentage of any Republican candidate since Bob Dole in 1996, when Ross Perot played spoiler for a second time. But I think we're still one more electoral cycle and/or a Hispanic President/VP candidate away from seeing Texas go blue. But polls on this one are all over the place, with many results being within margin-of-error. This could go either way, but given the numbers game, if Texas goes for Biden, it's likely a blowout anyways and Texas is just letting Biden hit 400+ EVs.
So let's take those three off the board and give them to Trump, and let's give Minnesota to Biden. Nevada is still close enough that it stays in play. The total is now 227-158.
Now, let's just be conservative as hell as give Trump any states where he's currently polling ahead. That takes Arkansas, Iowa and Georgia off the board as well. The one poll in Arkansas is 3 months old, Biden is losing ground in Iowa, and while I think Georgia will be very close, it also has a proven history of voter disenfranchisement and assorted shenanigans. If Georgia does go for Biden, that's a likely sign that Texas and Arkansas might actually be much closer. So now it's 227-186.
Arizona and Wisconsin are both over +5 for Biden. The polling data in Wisconsin has stayed quite stable since June. So let's be mildly optimistic and give those to Biden. The total is now 248-186. Biden only needs to pick up 22 electoral votes to win, Trump needs 84.
The remaining contests:
Florida (D+1.2, 29 EV, Trump)
Michigan (D+4.2, 16 EV, Trump)
Nevada (D+4, 6 EV, Clinton)
North Carolina (D+0.8, 15 EV, Trump)
Ohio (D+2.4, 18 EV, Trump)
Pennsylvania (D+4.3, 20 EV, Trump)
Biden only needs to win 2 out of these 6 (or just Florida). Trump needs to win 5 of 6, and Florida absolutely has to be one of them. (There is an extremely rare chance that Biden wins Nevada and North Carolina but loses the rest, which would make for a 269-269 tie and likely spell the death of the Republic as we know it). Mind you, Biden is currently leading in all of them, and he is widening the gap in Michigan and Pennsylvania.
If Biden were to win all of them, it would be a 352-186 win, somewhere in between Obama’s two win margins. If he were manage to outperform and take Georgia and Texas, it would be a 406-132 shellacking, on par with George H. W. Bush’s takedown of Dukakis in 1988.
If Trump were to actually win all six, it would still only be a 290-248 win, weaker than his 2016 win. And it may give him tremendous cause to regret all the noise about voter fraud, as any close state results would undoubtedly be hotly contested, and it would only take a pair of them being questionable (and/or Florida or Texas on their own) to invalidate the result.
TL;DR – There’s really only six “battleground” states, Biden is ahead in all of them at the moment, and only needs to win 2 of them (or Florida) to win.