Friday, September 30, 2016

Demos help Clinton nationwide, but may keep it closer in Wisconsin

While I'm not a huge believer in the smug "demographics are destiny" mentality that afflicts far too many in the Democratic Party, causing them to write off and lose too many rural areas and state legislatures, that doesn't mean there isn't some truth behind it. David Wasserman and Aaron Bycoffe came up with a great tool for FiveThirtyEight.com called the "Swing-O-matic" that explores voter demographics, and I encourage you to have some fun with it. Here’s their explanation of how this works.
We started with the results of the 2012 election and the support for each party’s candidate by the five demographic groups. We then adjusted the size of those groups based on four years of population change. When you adjust the vote and turnout above, our model recalculates the results for each state — as well as the Electoral College outcome and the national popular vote — taking into account how much of the state’s electorate the group accounts for.
So I wanted to see if Donald Trump’s “stir up bitter white guys” strategy could work well in propelling him into the White House. What I did is made the following changes from the 2012 results, based partially on polls I’ve seen, and partially on some conservative assumptions of what I think might happen.

Adjustments to 2012 electorate
College-educated white- Goes from 56% GOP to 52% Dem. Turnout stays the same at 77%.

Non-college whites- Goes from 62% GOP to 70% GOP. Turnout goes from 57% to 62%.

African-American- Stays at 93% Dem and 66% turnout.

Latino- Goes from 71% Dem to 80% Dem. Turnout goes from 48% to 50%.

Asian/Other Goes from 67% Dem to 70% Dem. Turnout goes from 49% to 50%.

Put that in the Wasserman and Bycoffe Swing-O-Matic, and the result is…

Clinton 347
Trump 191
Popular vote- Clinton 52.9%, Trump 45.4%

This would be a bigger win than Obama had over Romney in 2012, and the popular vote is virtually the same as Obama’s 9.5 million vote win over McCain in 2008.

What’s intriguing within this projection is that even though only one state changes from 2012 (North Carolina goes Dem), some close Dem states in 2012 become blowouts, while other Dem states actually get closer, including here in Wisconsin.

New “blowouts”
Virginia- From Dem +3.9 in 2012 to Dem +9.8
Florida- From Dem +0.9 to Dem +7.5
Nevada- From Dem +6.7 to Dem +10.4
Colorado- From Dem +5.4 to Dem +7.6

Closer blue states
Iowa- Dem +5.8 to Dem +1.1
Wis.- Dem +6.9 to Dem +4.3
N.H- Dem +5.4 to Dem +4.0
Ohio- Dem +3.0 to Dem +2.4

The reality of Wisconsin having a lot of low-educated whites helps explain why Trump was hanging out in Waukesha this week, and those demographics help explain why Trump’s best "Obama states" in polling over the last month have been Iowa and Ohio. We’ll see if that holds after this disastrous week, however, because a whole lot of women fall into that “non-college white” category, and I’m thinking they won’t take kindly to the Donald’s fat-shaming and piggish behavior. Particularly when the main-party opponent is the first female to be a major-party candidate.

I admit to being skeptical of the Florida and Nevada numbers being that easy of a win for the Dems, but demographics (and Trump’s racism) could be Clinton’s “secret weapon” in those states. And if she wins Florida, there’s pretty much no way she won't end up being president, if you do the electoral math. On a related note, a number of “red” states become pretty near purple under the scenario I created, including the biggest one.

Close red states
Georgia- GOP +7.8 to GOP +1.1
Texas- GOP +15.8 to GOP +3.5
Arizona- GOP +9.1 to GOP +3.6
S.C- GOP +10.5 to GOP +5.7

So how in the world can Trump win with this “bitter white guy” strategy? Basically it would be if no one else reacted to his race-baiting and misogyny, other than low-educated white folks. Here's a Swing-O-Matic combination that got Trump to a 282-256 win over Clinton, and Wisconsin would be among the states that flip his way in that scenario.

College-educated whites- Stay at 56% Republican. Turnout goes from 77% to 76%.

Non-college whites- Goes from 62% GOP to 70% GOP. Turnout goes from 57% to 64%.

African-American- Stays at 93% Dem and 66% turnout.
Latino-Stays at 71% Dem and 48% turnout.
Asian/Other Goes from 67% Dem to 68% Dem. Turnout stays at 49%.

I really wouldn’t bet on this outcome happening on November 8, and it shows how long the odds truly are for the GOP to win the presidency these days. In the real world this November, you’d have to think that Latino and Asian/Other turnout will go up and they'll vote Dem after the numerous insults that Drumpf has thrown at Mexicans and non-Christians (among many others). Throw in the fact that no one with an IQ over 80 and an ounce of decency will vote for the Orange one (meaning he loses the white, college-educated vote), and a more likely scenario is that this race breaks out for Hillary, and she ends up winning BIG.

Obviously, there's still a ways to go, and some kind of wacky event could change this. But even Nate Silver’s Trump-sympathetic model has turned hard in Hillary’s favor, as she has gone from a 55% chance of winning on Monday to 64% today. Not coincidentally, her lead keeps growing as post-debtate polls are being done and reported, and with Drumpf going on 4am Twitterstorms to personally attack former Ms. Universe contestants, I can’t see that trend reversing any time soon.

No comments:

Post a Comment