Tuesday, February 28, 2017

Politics As Religion

I'm catching up on my "Conversations with (Economist) Tyler Cowen" podcasts (earlier on Tyler). Here's a great quote from an interview with social psychologist Jonathan Haidt:
National politics is different from local. National politics, I believe, is much more like religion than local politics is. If you take it all the way down to the very local level — who the dogcatcher is, who the treasurer is of the town — that’s all very practical stuff. People are very worried about their property values and things like that. It’s not very ideological. National politics is much more like a religion. The president is the high priest of the American civil religion
And to be clear, I think this is mostly a negative.

Saturday, February 04, 2017

Why I Was Wrong About Trump's Victory (But Not Necessarily About His Presidency)

My brief streak of correctly predicting the presidency while never actually voting for the winner has ended. Donald Trump is now the 45th President of the US and few saw it coming (not even Trump). Last week I got to do a second presentation to my campus to update them on my own prediction failure and help them process why exactly the polls and national media were so wrong. Here was my explanation:

The Polls Weren't Wrong
When I first spoke to my school in late October, Clinton had about a 70% chance of winning in the polls. We all forgot the implication that 30% is not zero and far from it. He had a 1/3 chance of winning, which is still pretty possible. Secondly, the polls were off, but the 2016 polls were actually slightly more accurate than the 2012 ones. Only in 2012, the polling error didn't change the outcome. Here's an eerie title from Nate Silver (who I vowed to never link again) just 4 days before the election: "Trump Is Just A Normal Polling Error Behind Clinton". And finally, last minute undecided voters (who can't be measured well) did seem to shift towards Trump (was it Comey, Russia, Johnson, or likely Clinton's fault?).

Very Very Close Election
It looks now that Hillary Clinton is going to get more than 2.9 million votes than Donald Trump. In fact, if you convince 38,595 Trump voters in close states to switch to Clinton she wins. That's close. But of course the Electoral College system choses the president. That was relatively close too. Trump's victory is 46th out of 58 in past presidential elections.

No Obama 3rd Term 
The now well known "Prediction Professor" has a successful 13 yes or no question system for predicting a win (in this case, a Trump win). You can read them all here, but most of them come down to much larger factors beyond the candidates themselves. Very rarely does a political party get 3rd term in the presidency (it takes a Jefferson, Jackson, or Roosevelt) and it seems now it was Clinton's election to lose. If fact, I think Biden was the Dems only chance.

Party Loyalty Trumped (Literally)
The most surprising thing for me this season was just how little Trump's surprises mattered. Since his initial Birther claims in 2011 and every unelectable thing he did since, I'd assumed he couldn't win. I assumed the Republican base would not get behind someone who wasn't a Republican just a few years ago. I assumed... well you know what happens...

  • 81% of white evangelicals supported Trump, more than voted for last 3 Republicans (looks like they didn't read my post on the abortion)
  • Hispanics and African Americans voted for Trump more than the last Republican (at the end of the day, Hillary was no Obama)
  • Midwest/Rust Belt (PA, MI, WI, OH) voted with what they hoped for their wallet over any dreams of social justice
  • Rural America is underrepresented in our nationally culture, but purposefully overrepresented in the Electoral College, and voted for Trump by a huge margin 

You can see the rural/urban divide in the nation, my home state, and even in my precinct here in the city of Greenville
Trump won in at least part because he was Trump, and if there's one thing his first 2 weeks as President prove, he's still Trump.