Election Night: What States to Watch

So, it’s Election Day, and you want to know who the president-elect is before everyone else.
Well, as it happens, you could figure it out without looking aanywhere outside Virginia! Unless I’m wrong.

I’m basing this on two things: (1) Virginia has continued its slow drift from substantively more Republican than the nation (2004) to substantively more Democratic than the nation (2016), and (2) the effect of Northern Virginia is predictable enough that the Commonwealth’s result can be derived from the first 1% that comes in. If you want to use this Virginia Predictor Model, you’ll know who won the election fairly quickly.

First, though, let’s assume that can’t be done, and do a quick tour of the states that count.

7PM: New Hampshire. This is the first key state at the presidential level. It is the northeastern state most vulnerable to turning Red. Moreover, as it’s so small, it will be called early (when George W. Bush won the state in 2000, it was called in the first 45 minutes). If Clinton wins NH, it’s a good bet she has also held Pennsylvania (which has been a Republican tease for all of this young century) and Michigan. If Trump wins it, however, we’re in for a long night.
Virginia also closes at 7. See the Virginia Predictor Model below for more.

7:30: North Carolina and Ohio. If Clinton wins Ohio, this is over. Don’t expect it though. Carolina is more intriguing. It should be noted that the GOP statewide candidates have repeatedly underperformed in the Tarheel State: McCain in ’08, Romney in ’12, and Tillis in ’14 (of course, the latter two still won there). Given how close the polling is now, that and the HB2 backlash point to a very narrow Clinton win, which would helpful as insurance for the 8PM Watch-It states (see below). Because NC closes so early, it should fall to a candidate before midnight. That’s as far as I’ll go here.

8PM: Florida, Maine District 2, Michigan, and Pennsylvania. Odds are northern Maine will follow it’s neighbor NH – and that should be known before 8PM (see above). As for the latter two, the only reason Michigan is even being considered is due to Sanders’ primary win leading people to wonder if Clinton will underperform twice. I wouldn’t bet on it. The protectionist vote that dominated the Great Lakes State in the 20th century began to fade in the 21st. As for the Keystone State, as I’ve written above, it has teased Republicans before without result, and I don’t expect this year to be any different. If Trump does carry one of those two, however, Clinton will need both NC and Nevada. If Trump carries both, Clinton will need NC, NV, and Florida, which I see not revealing a winner for its 29 electoral votes until tomorrow…at the earliest. If Clinton wins Florida, she wins NC, NV, and the White House, but I’d be very surprised if she wins the Sunshine State while losing both MI and PA (losing only one is a very different matter).

9PM: Colorado. A Clinton “firewall” state. Not as likely to fall as PA or MI, IMHO. If Trump wins here, he’s probably already won at least one of those two, and the election as a whole.

10PM: Iowa and Nevada. Odds are Trump will win Iowa. Of course, we said that in January and look what happened. As for Nevada, it’s looking to me like a rerun of the Reid-Angle race in 2010, where Angle looked like the slight favorite but Reid had the votes. It’s small enough that the count should go quickly…I think. There is Vegas, after all.

Bottom line: if Clinton holds NH, PA, and MI, she wins. If she loses NH, she needs NV. If she loses PA or MI, she needs NV and NC. If she loses PA and MI, she needs NV, NC, and FL (which I don’t see that happening if she can’t win PA or MI), or she needs a concession speech.

Now, how does Virginia fit into all of this?

Well, if Virginia is truly drifting leftward relative to the rest of the nation, then we can expect Clinton to do better here than nationwide. We don’t know how much, though, until the votes are counted. We do know that northern Virginia (particularly Fairfax County) has become such a big part of the puzzle that Republicans need an initial advantage (defined as the margin with 1% of the vote in) of over 25 points to win the Commonwealth (Gillespie was exactly at +25 with 1%, and ended up -1). Odds are the polarization has grown since 2014, so it’s safe to say Trump probably needs to be at +30 with 1% in to win Virginia.

Of course, Trump doesn’t need to win VA, thus we need to know how much more “blue” our state is, and we don’t know that yet. So we have to use ranges. These are what I think will work.

If Trump is at 65% or higher in Virginia with 1% of the vote in, then odds are he has won Virginia, and that means quite a few other states in Clinton’s “firewall” (NH and CO, to say nothing of MI or PA) have fallen, and Trump will be President-elect.

If Trump is in the low 60s (60-64%) in Virginia with 1% of the vote in, then odds are he has lost Virginia, once Fairfax comes in as midnight approaches. Still it could be a very close result, boding well for NC and FL, and possibly NH. With NC, FL, NH, ME D-2, and NV, Trump could cobble together 270 electoral votes. It’s hardly a certainly, but it can’t be ruled out.

If Trump is under 60% in Virginia with 1% of the vote in, then he’s toast. Clinton will win Virginia by at least 7 points (and possibly double digits), and odds are she’ll win NC and FL while holding her entire “firewall.”

So, if you think my assumptions are sound, Virginia’s first 1% will give you all you need to know.

Then again, if you think my assumptions are sound, you haven’t been reading me very much.