States with new poll data added since the last update: All 50 states and DC.
Notable changes in: Florida (29), North Carolina (15), Wisconsin (10), Colorado (9), Nevada (6), New Hampshire (4), and the tipping point
It has been about two days since the last update. While there are mixed changes since then, on balance this is a good update for Trump, and the race looks closer than it did two days ago.
There will be one more update later today to capture any last minute polls released after the cutoff for this update. (There have already been a handful.)
Good for Trump
- The expected case moves from Clinton winning by 108 EV to Clinton winning by only 8 EV.
- The tipping point moves from Clinton by 1.8% in NC to Clinton by 1.6% in PA
Good for Clinton
- Trump’s best case has deteriorated from a 112 EV win to only a 66 EV win
- Clintons best case remains steady at a 210 EV win
The fact that Trump’s best case is getting worse is pretty much irrelevant. Either candidate winning all the close states is a very low probability event. A few months out it is good to show how things might reasonably move. But on election day, the expected electoral college and the tipping point are the two numbers to watch.
In the last update I speculated that because the tipping point and expected electoral college votes were going in different directions Trump may have topped out. But the movement toward Trump seemed to have a little bit of life in it yet. With the latest batch of update, the Election Graphs poll averages for Florida, North Carolina, and Nevada all move from just barely blue to just barely red. And the tipping point gets a bit narrower too.
Some have suggested there may a strategy to intentionally release Republican leaning polls at the last minute specifically to influence the poll averages, and the impact of some of these polls has been shown to be significant. This probably has something to do with the moves here too. But this site’s philosophy is to basically include all polls with very few exceptions, and when you throw everything in, this is the move you get. Given non-poll-related information on early vote results and such, I have my doubts about Florida and Nevada, but this site will stick to what the numbers say…
…and the numbers say that it is looking very close again. In the expected case Clinton wins by only 8 electoral votes. To flip the election, Trump only has to win Pennsylvania, and Clinton’s lead there is down to 1.6%.
The tipping point margin isn’t quite as small as it it was in September when it was down to 0.4%, but it is close enough that you could easily imagine a Trump win. Right?
Lets check the sites doing odds again for what they think the chances of a Trump win are…
- 28% – FiveThirtyEight
- 20% – Election Betting Odds
- 16% – New York Times
- 16% – HorsesAss
- 12% – PredictWise
- 8% – Daily Kos
- 2% – Huffington Post
- <1% – Princeton Election Consortium
The median estimate of Trump’s chances 2 days ago was 12%. Now it is 14%. So, slight increase, but not a huge one. And if you look at the individual estimates above, many actually went down since two days ago. But if everything is closer, why?
Well, each of the models of course have their own methods, and the movements in each would have a different explanation. Some of them discount or don’t include at all some of the pollsters showing the Trump movement that has moved my averages.
But just looking at the state of the race as a whole instead of specific models… Trump has never been ahead in the electoral college, by most measures he hasn’t ever led the popular vote either, and he’s also never led in Pennsylvania (or the other further off states he would have to flip to win).
So even taking the averages here at face value, and not doubting the status of Florida or Nevada, to win Trump would still need to break his September ceiling and do better than he has done at any time in this race so far. That is possible, but not likely.
Add to that the indications of high Hispanic turnout in several states, notably Florida and Nevada, and that some of the recent polls may be intentionally skewed… and it seems like if someone over performs their polls, it is more likely to be Clinton. But that is looking at things other than the polls themselves.
All and all, the new median of a 14% chance of a Trump upset seems about right.
Maine CD2, Pennsylvania, Nevada, North Carolina, Florida, and Iowa are all within 2% at the moment. They could still go either way based on the current averages. That is enough to swing the election.
So, as has been the case for awhile, a Trump win would be surprising, but should not be shocking. Trump has a path.
We shall all see soon enough.
In New Hampshire we already have some results of course… Dixville Notch, Harts Location, and Millsfield have already reported results. The totals: Trump 32, Clinton 23, plus a handful of votes for others. So there you go.
For real meaningful results, we now have less than 8 hours until parts of Indiana and Kentucky close. Then less than 9 hours until the first full states close and the states start getting called one way or the other.
We are almost there.
The following are the detailed charts for all the states that influenced the national changes outlined above.
Moving from Clinton to Trump
No longer possible pickups for Trump
Also impacting the tipping point
A note about faithless electors
In the past few days there have been reports of first one, then two, expected electors from Washington state saying that that have decided to or are considering being “faithless electors” by voting for someone other than who they are “supposed to” when the Electoral College votes in December.
Yes, they can do that. In Washington state there would be a $1000 fine if they did, but they can do it, and it would count.
There was also one other potential elector from Georgia who suggested they would be faithless back in August. That potential elector came under extreme pressure and resigned, to be replaced by someone who would presumably vote the “right” way.
That might well happen to these two as well. But faithless electors can and do happen every few election cycles. In the last 50 years there were faithless electors in 2004, 2000, 1988, 1976, 1972, and 1968. That is 6 out of 12 elections during that period. Half. It would not be surprising if there were one or more faithless electors in 2016, especially given the contentious nature of the race. They have never made a difference in the final outcome.
The Election Graphs estimates show what the results “should be” given who wins various states. They do not take into account statements by individual electors on their intentions. If a scenario arises where it is close enough for electors who have made statements like this to potentially make a difference in the outcome or throw the race to the House, we will of course make note of that. And once electors vote in December, we will show the final tallies, including any votes cast by faithless electors.
A note about McMullin
Right now the Real Clear Politics average in Utah has McMullin 12.4% behind the lead. The 538 average in Utah has him 9.4% behind the lead. Although McMullin was a lot closer than anyone expected, he never got closer than 5% to the lead in Utah in these averages and now appears to be fading. 538 still has him in second place, but it doesn’t seem likely at this point that he will win electoral votes.
For more information…
This post is an update based on the data on ElectionGraphs.com. Election Graphs tracks a state poll based estimate of the Electoral College. All of the charts and graphs seen in this post are from that site. Additional graphs, charts and raw data can be found there. Follow @ElectionGraphs on Twitter or like Election Graphs on Facebook to see announcements of updates or to join the conversation. For those interested in individual general election poll updates, follow @ElecCollPolls on Twitter for all the polls as they are added. If you find the information in these posts interesting or useful, please consider visiting the tip jar.