So much for all of my analysis and hunches. Trump won, and I am still in the process of figuring out why. Given the closeness of the election,with the split in the popular and electoral college, I am resistant to simple explanations of what went right for Trump while going wrong for Clinton.  I plan to take some time to think about it and wait a few weeks until all the votes are counted.

I apologize to readers if my confidence increased your shock and suffering on election night.  I wrote what I thought based on my understanding of past elections.  Clearly there was something going on that I didn’t understand, or didn’t want to understand. While I am well studied in polling techniques and the political science of elections and the economy, that analysis was not the primary source of my confidence which ended at about 85 percent likelihood of Clinton victory on election eve. The primary source was my confidence in the electorate to make rational decisions, not necessarily the decision I would want in every election, but rational ones nonetheless.  That’s why this is so hard for me to understand.  For now my confidence in the essential goodness of our system and our people is quite shaken.

A lot of weird stuff happened in this election that strains simple explanations.  How come there was so little ticket splitting with Senate results matching Presidential results in every single state for the first time in maybe a 100 years or more?  As the author of campaign finance reform, why didn’t Russ Feingold win in Wisconsin if this election was about populism?  If this election was all about race, then why was Obama holding 55 percent approval ratings and would have easily beaten Trump had he been on the ballot?  The article closest to my own current thinking is by Ezra Klein, entitled

Clinton’s loss can be explained—Trump’s win rewrites what we know about American politics.

I suggest reading the entire article at:

http://www.vox.com/policy-and-politics/2016/11/11/13578618/why-did-trump-win

A few words from his excellent piece:

I don’t have a model of the American people that accounts for electing someone like Trump. He’s done too many things, said too many things, tweeted too many things that would typically be disqualifying in American politics. Remember when Mitt Romney was mocked for his car elevator? Trump has a house covered in gold. Remember when John Kerry was assailed for supposedly insulting the military by the Swift Boat Veterans for Truth? Trump slandered war heroes and Gold Star parents despite getting repeated deferments from Vietnam. Remember when John McCain was dismissed for seeming ill-informed and out of touch amid the financial crisis? Trump doesn’t know how NATO works or what the nuclear triad is.

Trump’s victory is unnerving in a way nothing else in politics ever has been to me — it suggests there’s no bar, no floor, no you-must-be-this-decent to serve. I thought more of my country.

Polls show that in narrow ways, the voters saw what I saw — people did believe Trump unqualified, unkind, dishonest, indecent. It just didn’t matter.

To explain Trump’s competitiveness, if not his win, you have to search for truly primal appeals that overwhelmed all that — the power of partisan identity, the fear of Others, a dominant racial majority that rose in fury against the idea that it was becoming a political minority.

I hope Trump is a better man as president than he showed himself to be on the campaign trail. But I can’t confidently explain his win. In some ways, I don’t want to — I am scared of the conclusions it forces.

For now I’ll close with my own words from an email I sent to my brother the day after the election:

I don’t really have any sage words. I take a tiny bit of comfort that Clinton is winning, and is likely to extend her lead, in the popular vote. At least we know that a bare majority of our fellow citizens favored her over Trump. I still believe in our republic and that the voters will hold Trump and his party accountable in coming elections.

 

6:49 PM PST

Much closer than I expected.  No decisive calls in critical states that Clinton must win.  I’m most worried about Michigan.  Florida may go to Trump, but that’s a must win for him, not for Clinton.  I expect a long night.

6:58

Clinton now leading in VA.  NC exits look better to me than FL or MI.  But these exit numbers are super close.

7:32

I agree that Michigan is likely to be the key. Unless Clinton can win in NC or FL, which seems unlikely.  Michigan looks very close to me.

7:37

NC exits no longer looking good for Clinton. Exits just don’t mean that much tonight.

8:06

Senate will remain Republican.  A lot of Detroit votes to be counted in MI. The exits don’t look good in AZ or Iowa.  NC just called for Trump.  It makes MI absolutely critical unless she some how wins FL, which I don’t see.

8:18

Exits look pretty good for Clinton in WI and NV.  At MSNBC Cornacki (sp) just layer out a path where Trump needs to win one of those three.  That’s correct, but I think MI is the realistic place for him to break through.

8:23

Latest Wisconsin exits make it look more like lean Trump. Ruh Roh.

8:44

Wisconsin and Michigan could both go for Trump.  Clinton is an underdog now.  Perhaps not the huge underdog the NYT model says, but still an underdog.

9:18

Clinton is now behind in PA, along with WI and MI.  I just can’t see her winning all three. It would be a miracle. NYT gives her a 5 percent chance.  That sounds about right.

She is likely to win the popular vote as Trump leads by about 1.2 million votes, but Clinton is likely to run up a 2 to 3 million vote margin in California.

10:39

Feels weird waiting for the final calls of states which are all but certain to mean that Trump will be the next President.  I was wrong about the election and am very disappointed. Clinton may win the popular vote, but that is meaningless in terms of control of the U.S. Government. For some reason though, it matters to me.  I want to believe a majority of my fellow citizens preferred Clinton to Trump.

The speech I am most interested in hearing is from President Obama.  I hope he has some calming words for us in the coming days.

 

 

Polls today generally moved in Clinton’s favor.  Her average lead is closer to 4 than 3.  I think she is more likely to over-perform rather than underperform her polls. I think my electoral college predictions from yesterday evening are about right.  The next state she might get would be Ohio, if she indeed over-performs.

If FL or PA and NH are called before midnight eastern, we will likely have an official winner around 9:00 in the west. If it’s closer we may need to wait for CO and NM to be called.

Slate will have some real time estimates based on turnout at different precincts in swing states.  Unless all of the data from this new source is unambiguous, it’s not likely to be of much value in making an early call in a close election. I’ll check it out tomorrow.

Remember to vote if you haven’t already.

So today the FBI announced that the new emails do not impact their earlier conclusion not to recommend prosecution of Clinton.  I always believed this would be the case, I just didn’t know how long it would take the FBI to reach that conclusion.  However, the political damage was most likely already done by last Monday or Tuesday.  I don’t think this reversal will lead to a surge in Clinton’s favor.  It was a wasted 9 days with the focus away from Trump.  And that time away from Trump hurt Clinton’s margin, which might have been more robust but for Comeygate.

Depending which method for averaging polls that you elect to follow, Clinton leads national polling by 2.6 percent (Princeton Election Consortium [PEC] meta-margin of state polls), 2.7 percent (Fivethirtyeight), 2.3 percent (NYT Upshot), 5.2  percent (Huffpost Pollster), 3.9 percent (TPM Polltracker), or 2.2 percent (RealClearPolitics).  Most observers believe Clinton leads by 2 to 3 percent, with exception of Huffpost.  However, there is disagreement regarding the value of a lead in that range.  Nate Silver looks at historical polling errors which average 2 percent and posits that the final result could vary from a virtual tie to a substantial Clinton win.  Accordingly, he thinks Clinton’s chances of victory are about 66 percent.  Others like Sam Wang at PEC see a 2.5 percent lead as more robust, predicting a 99 percent chance of victory.  The NYT Upshot model is in-between at about 84 percent.  My personal hunch is similar to the NYT.

In more recent presidential elections since 2000, the error, if significant, has been in the Democrats favor.  However, in recent non-presidential years, it has been in the Republicans favor.  My hunch is the stronger Clinton ground game,  the fact that younger voters may be harder for pollsters to reach by cell, and that much of Trump’s strength is with less educated white voters (who may not be that serious about voting), suggest Clinton may outperform her final polls by just a bit. Of course I have no way of knowing for sure; that’s just what I think.  And my thinking is largely grounded in my earlier points that the fundamentals of this election favor continuation, rather than change.

For now I will stick with long held prediction that Clinton will win by between 4 and 7.6 percent.  My best guess is in the range of 4 to 5 percent.  I expect Clinton to win the 23 states listed in the Upshot Model totaling 268 electoral votes. In addition, I expect her to win the following swing states: FL, NC, and Nevada.  Probably New Hampshire as well, for a total 322 electoral votes.

I would note that if my hunch is wrong, and the Clinton margin is just 1 or even 2 percentage points, there is a risk that Trump may win the electoral college while losing the popular vote.  One good sign for Clinton is her strength in early voting in Nevada.  This could offset a loss in New Hampshire.  For Trump to win, he must take FL, PA, and Ohio.  Doing so enables his victory but does not guarantee it.  He needs one or two more blue states.

I’ll post again tomorrow night if I have any more insights. I want this election to be over.

No, not really.

Per my post on Friday, I tend to look at elections from a more fundamental point of view. Is the economy healthy, is the nation at peace?  If the fundamentals are strong, the electorate will tend to stay the course and return the incumbent party to office as in 1972, 1984, 1988, 1996, 2012.  If there are problems, such as in 1980, 1992, or 2008, they vote for change.  Campaign tactics and candidate likability matter in close elections, but tend not to overcome fundamentals.  The fundamentals favor Clinton.

Comeygate is not really a new story.  It’s just more about emails.  The media are much more worked up about emails than the average voter.  There is no reason to think that these emails contain a smoking gun indicating that Hilary compromised national security.  In fact, they may not even be new emails. Secretary Clinton was presumably the sender or receiver of these emails from a computer used by her aide Huma Abedin, and the FBI would already have them from their earlier investigation of the Clinton’s server.

That said, the timing is terrible for HRC.  It takes attention off of Trump and puts it back on Clinton.  Before Comeygate, there was a real possibility of a Trump collapse and a Clinton win in the 8 to 10 percentage point margin.  When it becomes obvious that one candidate is going to lose, their supporters lend to lose enthusiasm and may not turn out.  Comeygate re-engergizes Trump’s base and makes it more likely they will turn out.  I’ve always expected the election result to be between the two Obama margins.  Now I think we are much more likely to be at the low end of that range, a 4 percent win.

I don’t obsess over tracking polls. They tend to measure enthusiasm at the margin.  When a candidate is having a bad news day, less of their voters answer the pollster’s call.  I don’t really believe in wild polling swings after the convention bounces have settled out.  Although I am a fan of Nate Silver’s fivethirtyeight, I also appreciate the work of Professor Sam Wang at Princeton Election Consortium.  Wang uses a model based only on state polls, which he then averages into a “meta margin.”  Since May, that margin has been averaged around 3.5 percent in Clinton’s favor. Sometimes it move higher when news is good for Clinton (post convention) or  lower when she is having bad news. Right now the meta margin is at about 3.2 percent, suggesting an electoral college win of 317 to 221 according to Professor Wang.

Bottom line, I expect Clinton to win by about 4 percent, unless Comey is discredited and/or there is more negative news on Trump.  Both of these are real possibilities. The media understands that Trump is a much greater danger to the security of the union.  The public understands the strengths and weaknesses of both candidates and will make a rational decision.

Kevin Drum has a nice piece on Sam Wang’s work and a related discussion by the folks at the YouGov poll, which tries to deal with response rate changes by using weights based on the 2012 election.  Here’s a link:

http://m.motherjones.com/kevin-drum/2016/11/here-are-my-nominees-phrase-year

Kevin’s final paragraph mirrors my own thinking:

Clinton could easily lose another point before Election Day, or she could revert back to 3.5 and stay there. I’d bet on reversion to the mean. This election features two candidates who have been around a long time and are both very well known. Almost everybody made up their minds pretty early, and nothing much has changed for the past 12 months. Hillary Clinton will most likely win by 3-4 percentage points, plus maybe a little extra because she has a way better ground game.