Brief Musings on Polls

The polls have bugged me this year.  Back in 2016 Trump won the election with 63 million votes.  At the time he was a brash adulterer who left broken businesses, broken marriages, and broken promises in his wake.  He pitched his ultra-liberal sister as a potential Supreme Court Justice, and divided voters over a wall.  Yet 63 million people voted for him.  I will consider those as most likely locked in voters.  If someone voted for Trump in 2016 they are probably going to vote for him again in 2020.

The polls have Biden ahead by almost 10 pts.  If that were going to be true then Biden would have to get 77 million votes (assuming no third party votes).  Obama got 69 million votes, but the electorate has grown since then.  Maybe the growth will account for that shift.  Actually, that is possible.  If the 2016 election matched the turnout of 2008 there would have been 142 million votes.  In 2008 1.1 million votes went to third parties, so Joe Biden can match hit his target assuming he attracts as much of the electorate as Mr. Hope and Change, our first African American nominee of a major party, a generational candidate.  You can call Biden many things, but the face of a new generation he is not.  I find it doubtful that he attracts as much of the electorate as Obama did in 2008.

This assumes that Trump gets 63 million votes.  The reality is that a lot of Republicans in 2016 (me included) refused to vote for Trump.  I wrote in.  However, this time around I will vote for Trump. I’m guessing a lot of Johnson’s and McMullin’s support shifts back to Trump pushing him up to 66 or 67 million votes.

Well, it’s all about turnout, and the people who sat out in 2016 will show up this year.

Maybe.  However, nationally the turnout was 55.7%.  Not great, but not terrible (roughly the same as 2012).  The interesting thing is when you look at the swing states.

  • Michigan: 63% turnout (66.2% in 2008)
  • Wisconsin: 67.34% (69.2% in 2008)
  • Pennsylvania: 60.3% (63.7% in 2008)
  • North Carolina 68.98% (63% in 2008)
  • Ohio 71.3% (67.3% in 2008)

The reality is in states that decided the election turnout was high.  In the states with a dip the dip was smaller than Johnson’s vote tally.  If Biden returns to Obama’s turnout then Trump merely has to bring home Johnson/McMullin voters.

All together my issue is this – the polls just don’t make sense.  There is no way that Biden can win by 10 points.  That just isn’t in the cards.  Trump’s vote tally should go up this year since his base loves him, his original voters probably won’t have their minds changed, and he has convinced a few people (like me) to vote for him this year.  The pollsters know all that information, so what are they doing?

The world has gone mad.