2016 Election Post Mortem

 Thoughts on Trump Becoming President-Elect

I  called 46 states correctly (92%) and four states narrowly wrong (North Carolina, Pennsylvania, Florida and Wisconsin).  I said the vote would be Trump 46% (he got 47.5%) and Clinton 49% (she got 47.7%), both within the margin of error of polling.

There were two overwhelming factors that explained why Trump won.  Both of these factors were missed by the pundits and underestimated by me.

Wives Should Always Be Lovers Too

Women were expected to vote for Hillary Clinton.  Minority women and college educated women did so.  But white, non-college women voted 62% for Trump and 34% for Clinton.  These are the (often working) women married to the (often underemployed) blue collar men who went to Trump rallies.  Take a look at this link:

http://www.cnn.com/2016/11/09/politics/clinton-loss-women-politics/index.html#

Throughout the primaries and main campaign, Trump had a ceiling of 44% of the vote.  I thought I saw a three point margin between Trump and Clinton, so to her projected 49% I gave Trump 46% (a number higher than he had been polling).  I very strongly suspect that white, high-school educated women were, as a group, infuriated by the decision the Friday before the election by the FBI to drop further investigation into Clinton’s emails.  They couldn’t get away with breaking the rules like that and weren’t charmed by this power play.  Basic common sense drove them to Trump over the weekend and this was too close to the election for multi-day polling to catch it.

Reverse Coattails

Trump was cheap about the ground game.  He didn’t put forward the money to organize a get-out-the-vote drive nationally.  He had only a skeleton staff in key states.  This is a big part of why so many professionals thought he would lose.  Personally, let’s take a look at the four states I myself called wrong: North Carolina, Pennsylvania, Florida and Wisconsin.  I wrongly called all four of these states for Hillary.  All of these states had incumbent Republican U.S. Senators who were up for re-election; all got re-elected.  I don’t think it is too much to suggest that the Trump campaign rode the coattails of these senate contests.

Normally the presidential candidate leads in votes.  But Trump’s campaign was miserly and his smart mouth hurt him with the public over and over.  A lot of emphasis by the party, especially the Republican National Committee, was on a major attempt to maintain control of the U.S. Senate.   This was successful and carried Trump along.  President-elect Trump owes huge favors to the senators who won re-election in 2016, especially in the battleground states that put him over the top in the electoral college.

Populism?

Trump did run a populist campaign and he had huge crowds at his rallies.  His message was received enthusiastically and often broadcast live on television.  How did this effect the election?  In my opinion, it didn’t.  Turnout in 2012 was approximately 130 million voters, significantly above the 2016 level.  Romney received 47.2% of the vote in 2012, compared to Trump’s 47.5% of the 2016 vote.  Trump narrowly lost the popular vote.  Therefore the rallies and Trump’s populist stances are nearly specious.  The big effect of the rallies was to increase television exposure for Trump that allowed his base to grow to the size of Romney supporters four years earlier.  At no time during the 2016 election was Trump in control of his own destiny.  He didn’t control the RNC, he didn’t coordinate closely with the senate campaigns in the battleground states, and he didn’t have a plan to pursue the women voters.  Trump’s populism consisted of saying catchy things on television as an experienced game show host.

It makes more sense to say that President Obama’s leaning on the F.B.I. to drop an investigation of Clinton emails infuriated high-school educated white women during the last weekend of the campaign and gave Trump an electoral college victory and a near-tie in popular vote.  Most polls “roll” for three or four days and thus were not able to fully catch the political impact of the F.B.I. decision on Friday, November 4^th  by election day on November 8^th.

Because polling is partially retroactive, it couldn’t catch this last minute surge of Trump support above his normal 44% ceiling.  Most polls were honest and within the appropriate margin of error.