Friday, November 18, 2016

Polling predictions versus actual electoral outcomes in the 2016 presidential election, by state

The following table and map show Trump's electoral performance relative to RCP's two-way polling averages, by state.

The poll results were averaged except for the few cases where no polling had been conducted since October 1--in those states, the most recent poll was used as the RCP 'average'. There was no polling data available at all in four states. They are colored in black on the map. The rest of the states are colored according to how greatly Trump exceeded/fell short of polling expectations, the darker the greater his over-/under-performance. The five states where the RCP average was within half a percentage point of the actual results are shaded in an off-white color.

StateO/(U)
West Virginia24.2
Oklahoma21.4
South Dakota19.1
Tennessee16.2
Kentucky12.8
Alaska12.2
Maryland10.8
South Carolina10.1
Missouri9.6
Delaware9.5
Indiana8.6
Minnesota8.0
Wisconsin7.5
Iowa6.6
Ohio6.4
Louisiana5.7
Montana5.5
Rhode Island5.4
Arkansas5.1
Kansas4.5
Michigan3.9
Utah3.7
Mississippi3.5
Pennsylvania3.3
North Carolina3.0
Connecticut2.8
Maine2.8
Idaho2.6
Massachusetts2.1
Georgia0.9
Florida0.9
Virginia0.4
Colorado0.1
Arizona0.1
New Hampshire0.0
Vermont(0.5)
Nebraska(0.7)
New York(0.8)
New Jersey(1.8)
Texas(2.4)
Oregon(2.6)
Nevada(3.2)
New Mexico(3.3)
Washington(3.3)
Illinois(4.5)
California(6.5)
No data for
AL/HI/ND/WY



One thing that immediately jumps out is the tendency for the winner of non-competitive states to significantly exceed polling expectations (ie Trump doing even better than the lopsided polls suggested in dark red states like West Virginia and Oklahoma and Clinton doing better than the polls predicted in dark blue states like California and Illinois). The correlation between the margin of Trump's victory/defeat and his over-/under-performance relative to the polls is an impressive .63 (p-value = .0000002).

Another thing that astute commentators noted from the beginning but that the polling agencies never fully picked up on was the nationwide east-west divide. The dynamics in play here are the rootedness, manufacturing, and populism of the east versus the transplanted, service-oriented, libertarianism of the west. Trump did better than expected in the east and worse than expected in the west.

A few comments on where the polls went especially wrong:

- They drastically underestimated how the subjects of Hillbilly Elegy would come out in droves for Trump. These are the forgotten Americans he spoke about again and again during the campaign. They were as forgotten by the pollsters as they are by the political class.

- Trump was, save for Chicago Illinois, significantly underrated across the entire Upper Midwest and surrounding regions. Wisconsin, Minnesota, Michigan, Iowa, Indiana, Ohio, South Dakota--the polls didn't just miss in these states, they missed badly.

- Hispanics--illegal or otherwise--responded to the threat that the reconquista could be coming to an end by shaking off a bit of their apathy to vote at higher-than-expected rates.

Parenthetically, the exit polls should be viewed with the same skepticism that the pre-election polls were correctly viewed with. It's highly unlikely that if blacks dropped from 13% of the electorate in 2012 to 12% in 2016 that the percentage of voters with no more than a high school education dropped from 24% in 2012 to just 18% in 2016, even though that's what the exit polls--which initially indicated a comfortable Clinton win before being sloppily adjusted in light of the actual results--show. We'll have to wait for the Census and eventually the GSS for more precise and reliable data.

17 comments:

ADog said...

Trump was, save for Chicago Illinois, significantly underrated across the entire Upper Midwest and surrounding regions. Wisconsin, Minnesota, Michigan, Iowa, Indiana, Ohio, South Dakota--the polls didn't just miss in these states, they missed badly.

This is the "how" of the election. It remains to be seen if this signals a new alignment in the electoral map or if it was specific to Trump himself.

Dan said...

The best answer I have for the systemic polling error is enthusiasm, which was reflected in Trump's enormous rallies.

Trump supports would have been willing to crawl across a field of broken glass (or as I observed in a couple of instances while doing calling into swing states, arise from being bedridden, glory be, to cast their vote.

Me, calling North Carolina: "Hi, how are you doing today?"

Elderly lady on the other end: "Bad."

Me: Can you make it to the polling station to support Donald Trump?

Elderly lady on the other end, in a southern drawl: "I have so many health problems, I haven't been able to leave the house, but honey I'm gonna try."

Me, trying not choking up: We need you. God bless you.

Dan said...

The other reason for the poll gap may be that Trump had a powerful surge at the end. This is plausible:

(1) The Comey stuff really hit hard.

(2) Trump dialed up ad spending like crazy crossing the finish line.

(3) Trump campaigned like crazy into the home stretch. On Sunday 11/6, he did 6 rallies in 6 different states.

(4) Never Trumpers got on the Trump train in the end.

The way to see if there was a last minute surge is to see if he outperformed in states where most people vote on election day vs. states where most vote early.


Audacious Epigone said...

Dan,

Another way to check out the last minute surge is simply to look at state exit polling. Trump did much better among those who said they'd decided who they were going to vote for in the last week in PA, WI, MI, and MN than he did in the country as a whole on the same question, on the order of 15 points better.

Sid said...

I wouldn't underestimate the effect of the "Shy Trump Voter" in California:

1. When you live in a deep blue state, it's hard to work yourself up into voting for a Republican candidate. There's no way he's going to win - why bother?

2. Once you vote for Trump, you will have to lie to everyone but your closest family members and friends about your vote, and if you're not naturally Machiavellian, it's no fun to lie about something like that.

3. The MSM hardly covered it, but there was plenty of political violence targeting Trump supporters in California. Again, why bother voting? Even if no one finds out you voted for Trump, it's still scary to have to wonder if the mobs waving Mexican flags will clobber you and throw eggs at your wife.

4. As Steve Sailer noted recently, Trump won on the issues. Trump's brash, aggressive, nouveau riche personality was both an asset and a liability, and many Republican voters never came to like him. That is especially true in California, where people put huge stock in being "nice" and modest, avoiding blunt talk, and trying to appear that they don't really care about money. In short, Californians are much nicer and phonier than New Yorkers, and Trump is almost a caricature of a New Yorker.

When you don't like the candidate you agree with deep-down, it's easy to invent all sorts of reasons why he's equally bad as his opponent. Sure, Hillary may have defrauded relief aid to Haiti so she could dole it out to her donors and cronies... But Trump said one of his models was fat in the 90s! Oh the humanity!

VFM #7634 said...

4. As Steve Sailer noted recently, Trump won on the issues. Trump's brash, aggressive, nouveau riche personality was both an asset and a liability, and many Republican voters never came to like him. That is especially true in California, where people put huge stock in being "nice" and modest, avoiding blunt talk, and trying to appear that they don't really care about money. In short, Californians are much nicer and phonier than New Yorkers, and Trump is almost a caricature of a New Yorker.

I would've thought this a reasonable theory until Trump grabbed Wisconsin and made a serious run for it in Minnesota.

Audacious Epigone said...

Sid,

Right, what VFM said. "Minnesota nice" is a phrase for a reason. The cuck corridor is full of nice Flanders types, but they don't have the same sense of haughty moral superiority that exists on the West Coast. Moral posturing means a lot to coastal whites. The inland whites in California are more like the cuck corridor whites though.

Sid said...

Right, but that's why I listed three other reasons why Californian Republicans didn't turn out for him. The biggest difference between California and Minnesota/Wisconsin is that the former is plurality white, whereas the other two are notoriously non-vibrant. As such, Trump always had a shot in Minnesota/Wisconsin, but none at all in California.

Sid said...

Midwest cucks are genuinely nice people. Mormons in particular are almost creepy for how sincere they are. The problem with them is that they're so naïve that they think everyone else is as good natured as they are, so it often leads to disastrous policy implications. ("Yes, let's have Somalis move on in! Iraq is primed and ready to be a liberal, parliamentary democracy!")

In contrast, West Coast liberals are "nice," but more in the sense that they don't want any open disagreement with their doctrines. They will kindly ask you about your views on transgender issues, and then will angrily tell you to shut up when you answer honestly because it's "offensive" to the group norms. It's really quite phony.

Southern Man said...

Pretty amazing that Oklahoma - that reddest of red states, that didn't vote a single county for Obama or Clinton in all three elections - is second on that list. Did they really believe those numbers?

Audacious Epigone said...

Sid,

In contrast, West Coast liberals are "nice," but more in the sense that they don't want any open disagreement with their doctrines. They will kindly ask you about your views on transgender issues, and then will angrily tell you to shut up when you answer honestly because it's "offensive" to the group norms. It's really quite phony.

I lived in Seattle for a few years and have visited multiple times since then. That's a spot-on description. West Coast niceness is qualitatively less sincere than the genuine cuckery of the Midwest. It's correspondingly less neighborly. The point is well taken about how California never had a chance, of course--for a state where non-Hispanic whites are a minority, there's no way for a Republican to win unless whites vote more monolithically than non-whites do, which is a long way from happening if it ever does.

Southern Man,

It's as though Oklahoma is just an extension of Texas, and Texas was relatively close. Oklahoma is like Texas without the NAMs, though, and that's a crucial difference. The pollsters were talking about states like Utah and Missouri being in play, too. Ignorance, subterfuge, or a combination of both?

JayMan said...

Great work!

It would be interesting to add the date of the last available poll for each state.

Mil-Tech Bard said...

>> It remains to be seen if this signals a new alignment in the electoral map or if it was specific to Trump himself.

We will find out in 12 years.

The pattern of working class white males and their wives voting outside the South exactly as they do in the South is "realignment", if it sticks.

Reg Cæsar said...

Check out the "driftless area". Three Dems have lost that in Obama's lifetime: McGovern, Mondale… and Hillary. Now that's notable.

Audacious Epigone said...

Reg Caesar,

The Driftless Design or the Driftless Objective? The Sailer Strategy has a better ring to it, but there's a lot of overlap in what they require for success.

Reg Cæsar said...

Here's an analysis of the region's vote from 2012:

http://www.themoneyillusion.com/?p=17616

Audacious Epigone said...

Reg,

"Only God (or Nate Silver) knows the answer."

Ha, I wouldn't count on Silver knowing the answer!