Two things.
Will We Know the Results on Election day?
And the answer is... maybe. If Biden wins.
The reason for this is the states that allow mail in votes to be counted early. Florida, Colorado, Arizona, Ohio, Texas, Nevada, and North Carolina do. By varying amounts. There may be more, but these are the ones that are remotely competitive.
We're likely to see results from Nevada, Arizona, Colorado, and Florida because they give you a while before election day.. Texas gives a 4 day jump, but is really big so probably not. North Carolina is possible since I believe they have two weeks. Ohio is still determining when to start so, it's iffy.
Of those states, Florida, Ohio, and Texas are must wins. North Carolina and Arizona are not must wins themselves, but a Biden win in both would make a win virtually impossible for Trump.
So, if Florida is called for Biden on election day/night it would be pretty safe to assume that Biden is going to win. Similarly if Ohio is called for Biden, he almost certainly will win in the rest of the rust belt. If Texas is called for Biden, obviously that's checkmate. If North Carolina is called for Biden, we can be pretty certain of the result... If Arizona and NC are both called, then we can be confident.
It's also not impossible for Biden to outright win on election night, but it's very unlikely. If all of those states except Texas are called for Biden, that along with the safe states would give Biden 275. If Texas is called for Biden, he can get to 313. So, in theory Biden can secure the win on election night, even with 100 votes left to be counted.
Could Trump actually win on election night? No.
If all of those states except Colorado voted Trump, he'd be at 254. If Trump got all of those and Colorado, he'd be at 263, but that seems really unlikely. Barring something really bizarre like Trump getting 100% of Pennsylvania's in person vote, there's nowhere else to get that last seven votes.
But, we could potentially see signs that would be very troubling to the point where we could consider Trump a practical lock. For example, the above scenario of Colorado and Nevada voting Trump. If we had enough data in Ohio to project that Trump would win by an overwhelming margin (like 12ish), that would also mean Biden would almost definitely lost. If Trump won by like 7 in Florida, that wouldn't really guarantee it, but it would certainly be a big scare. There could be warning signs but not a death knell.
We should probably brace for a tumultuous week long (or maybe more) election "night". But... we could get lucky (at least if you support Biden) and have a result in the wee hours of November 4th.
And one more thing regarding 2016 since Biden brought it up in an attempt to scare me into giving him more money...
Some people are still convinced by the polls showing that Trump is outperforming Biden compared to how he was doing against Hillary in battleground states, mostly based on RCP's average. RCP had Clinton up 4.8 compared to 4.6 for Biden. However, in the national polls realclear has Biden up by 9.7% as opposed to 5.5%.
First off it's worth pointing out that this is Clinton's peak. After the debates she shot up, and then things closed up in the last two weeks. Biden's lead is consistent so it's not as likely that will happen again.
More to the point, how do we explain that difference between the national average and battlegrounds?
If people think the battlegrounds are still underrating Trump, those numbers are hard to reconcile. If Biden is outperforming so much in the national average, why would this small set of states be immune to the changes?
The most likely explanation is that the polls are better. If Biden indeed winds up performing this well on election date, outperforming Clinton by 5-7 points, then we would expect him to win Florida by 3-5, Pennsylvania by 5-7, be competitive in Ohio and so on. If the race shifts that much than states like Ohio, Penn, and Florida which are typically close to the national average should see a relative swing.
As for the other possible explanations... well I can't really think of any good ones. It could be that the demographics shifts. For instance, Arizona is shifting democratic at a faster rate than the rest of the country. So it could be the rust belt states are shifting in the opposite direction of the country making the results still off. But that doesn't seem likely.
Basically, if the national average was correct, what we would expect to see in the swing states is exactly what we're seeing. Conversely, if the polling in the swing states is right, then the national average is about what we'd expect. So, they help to verify each other. The only other real explanation is that they're both screwed up and to the same degree in the same direction.
Last edited by JWeinCom - on 08 October 2020