Here’s another post on how to go to bed early on election night satisfied that you know the outcome. Of course this won’t work if the election is really close but if one candidate has a solid lead it should be possible.
I’m basing this largely on the current projections by fivethirtyeight.com. Currently they give Clinton about a 65% chance of winning–good odds but hardly a lock. Other sites give her a better chance. I’m going with fivethirtyeight because their record in past elections has been very good and their methodology makes sense to me.
[UPDATE: Fivethirtyeight’s final forecast based on the last polls released before the election gives Clinton a 71.6% chance.]
Digging though the numbers it appears that Trump has a list of states that he must win or he’s toast. These include Arizona, Florida, Iowa, North Carolina, Ohio, and probably Nevada.
If he gets all of these Clinton can still win if she holds on to her “firewall states”: Colorado, Michigan, New Hampshire, Pennsylvania, Virginia and Wisconsin. If she loses any of these while Trump wins his “must haves” then Trump will probably win.
So let’s look at states from these lists that are in the eastern half of the country so results may be available fairly early.
States where the polls close at 7 PM Eastern
Florida. An absolute must-win for Trump. If he is clearly losing it you can go to bed. It’s over.
New Hampshire. Bad news for Clinton if she loses it, but only if Trump gets all of his “must haves.” Even then she might be able to make it up elsewhere.
Virgina. Clinton should win this fairly easily. If not it’s probably a sign that she’s running well behind her polls nationally and is in big trouble.
States where the polls close at 7:30 PM Eastern
North Carolina. A must-win for Trump.
Ohio. A must-win for Trump.
States where the polls close at 8 PM Eastern
Pennsylvania. A big Clinton firewall state. If she loses this she is likely in trouble.
Expect a long night if…
Several hours after the polls close Clinton has most of her firewall states and Trump has most of his must-have states but the others are too close to call.
Bad scenario #1
Fivethirtyeight give about a 12% probability that Clinton wins the popular vote but loses the Electoral College. (2000 revisited.)
Fivethirtyeight gives about a 1% chance that nobody gets a majority in the Electoral College. In that case the President will be chosen by the House of Representatives, almost certainly controlled by the Republicans.
Even Worse Scenario
Two members of the Democratic Party Electoral College slate in the state of Washington have suggested that they won’t vote for Clinton. They are disgruntled Sanders supporters so they won’t be voting for Trump. Still in a close enough race that could still throw the race to Trump, or to the House of Representatives.
Since they have publicly stated their intentions, my guess is that they will be replaced before the Electoral College votes. My real worry is that there may be other disgruntled Electors who are keeping their mouths shut. The public will probably grumble but accept Bad Scenarios #1 and 2 above, but if the election is decided by “faithless electors” then we will have the worst constitutional crisis since the Civil War.
[UPDATE: A final possible leading indicator:
Exit poll results are supposed to be kept secret until the polls close, but it seems likely that the results will be leaked early to Wall Street insiders. In this election the stock market has tended to go up when Clinton had a big lead in the polls and to go down when Trump’s chances rose. So if this afternoon the S&P 500 Index is up sharply, take it as a sign that the exit polls show Clinton winning. If it is down sharply assume that Trump is doing better than expected.
Caveat: in the past exit polls have often given misleading results due to difficulties in getting a representative sample, shy voter effects and the systematic exclusion of third-party candidates.]