Sunday, November 6, 2016

4th General Election Forecast: Clinton 49.7, Trump 44.6, Johnson 4.1, McMullin 0.8, Stein 0.7 (Electoral College 322, Trump 216)

In about 48 hours, we will know the identity of the election winner. That winner will be Hillary Clinton. The margin will be similar to the election of 2012 but there are a handful of states that could go either way.  Trump would need to win every single one of them to get to 270 votes.  If you go to Predictit.com and buy shares in the states most likely to vote for Hillary, you can get 274 EV and the cheapest state would be Pennsylvania at 79 cents. Even if Trump pulls one truly significant upset in one of those states, Hillary could save the election by winning one of the other states where she is favored. Florida (69 cents) or North Carolina (61 cents) plus New Hampshire (76 cents).

But the elections could be close. If Nate Silver again correctly predicts every single contest this time, and his model does not change tomorrow, she will win 278 to 260. But consider that this includes Trump winning Nevada, because Nate's models, unlike betting markets, do not factor in the early vote. I think Hillary stands an excellent change of exceeding 300 Electoral Votes, but that's not guaranteed.

Trump's route to 270.

To build Trump's path to 270, one would expect to start him with the 206 Electoral Votes that Romney won in 2012. The day before the 2012 election I wrote a blog post called Romney has 5 Paths to 270 (and one to 269.)  This was based on my reading of the polls, the futures markets and the 2008 vote.  For each of those paths, I assumed that Romney would win flip North Carolina because it was his best chance of a pick up. He did flip NC back to red, but four years later it is the state most likely to flip back to blue.  So let's start Trump with 191 EV.  The only other Romney states he has a significant chance of losing are Arizona, Nebraska's 2nd congressional District and possibly Utah. But a close election would mean he turned out enough of his voters to hang on to the shaky red states. So he starts with 191. From that starting point he needs to add:

Three  Contests where he is favored:
1. Maine's 2nd congressional district, where he is currently the slight favorite. (192)
2. Iowa, where polls have shown him ahead for awhile. (198)
3. Ohio, where his anti-free trade positions and his support for the coal industry have resonated. (216)

These are the only contested states where he is the betting favorite.  I think he is likely to win all three of them, but I expect Ohio to be close.

Nothing Could Be More Floridian.
4. Not unlike Romney in the previous cycle, he almost certainly wins NC if he wins the presidency. But unlike last time, the polls and betting markets are against him, slightly. (231)
5. Florida, Florida, Florida.  I once thought FL would be an easier win for Trump than NC but today's early vote numbers looked great for Democrats. I really do favor Hillary here, but it's easy to go broke betting on the good sense of Floridians. (260)

The Last 10 Are the Hardest.
For weeks I assumed Donald's best path to 270 would involve Nevada and New Hampshire getting him to exactly 270. The NV numbers do not look good for Republicans and Hillary is also a heavy favorite in New Hampshire, which seems to be trending leftward as the Boston suburbs sprawl ever northward. This is still his best path, and you go to election night with the polls you have, not the polls you want. (270)

A Big 10 Upset.

If Trump fails to run the states listed above, he will need to pull a significant upset elsewhere. His best chances are in Big 10 Country:

A. If Trump wins Pennsylania (20 EV)  or Michigan (17 EV), he could win even if he lost North Carolina and New Hampshire. He also could win the election if he kept NC but lost both NV and NH.
Hillary's price on Predictit.com is 83 cents in PA and 79 in MI.

B.  If Trump wins either Wisconsin or Minnesota (each with 10 EV), he could afford to love NV and NH. Hillary's price on Predictit.com is 86 cents in WI and 88 in MN.

When the first Comey letter came out 10 days ago, Predictwise.com gave Hillary a 91% of winning. That got down to 82% before starting to rebound. She is nearly back to where she was, standing at 89%. Trump has a tough road to 270. With a good result in NC and FL, the election may come to a merciful end a couple hours sooner than expected.


ForecastPopularElectoralChanges
First (July 10th)Clinton 50, Trump 44, Johnson 4, Stein 2.Clinton 348-190(From 2012) Clinton wins NC and NE-2
Second (Sep 26th)Clinton 49, Trump 45, Johnson 4, Stein 2Clinton 340-198Trump wins IA, NE-2 & ME-2
Third (October 16th)Clinton 50, Trump 43, Johnson 4, Stein 2, McMullin 1Clinton 359-179Clinton wins IA, NE-2, ME-2 & Arizona
Fourth (November 7th)Clinton 50, Trump 44, Johnson 4, Stein 1, McMullin 1Clinton 322-216Trump wins AZ, IA, OH, NE-2, ME-2









No comments:

Post a Comment