How does this happen? A futile explanation.

"PotbellyJoe and 42 others" (potbellyjoe) 11/09/2016 at 11:42 • Filed to: WTFlock, Politics

15

28

My credentials: I have a decade of marketing research under my belt and currently work as a data analyst for a huge non-profit that you have all heard of and have an opinion about. During my time in MR, I worked with major MNCs for products that were in everyone’s home. Be it pasta, TVs, mobile phones, women’s health products, or alcohol to name a few. I will try to tie all of these paths together as I attempt to deconstruct what occurred last night. I don’t know it all, or even a lot, but I can read numbers.

1. Why were the polls wrong? Or were they?

There is a very strange dynamic to the candidates this year. On one hand you had a well-known, divisive candidate who has been in the public eye for 30+ years. Her name has been mentioned in every presidential race since her husband left office as a potential candidate and even while in office, her political ambitions were well-known. There is no one in America who does not have a formed opinion of Hillary Clinton. On the other hand, you had a cult of personality, wrapped in some of the most despicable and tumultuous primaries and social misbehavior imaginable for a front-runner. It was a perfect storm, but for what?

There is a well-known issue in opinion and behavior research that shows a bias towards social compliance and social norms. When someone knows their answer is not in line or compliant with expectations, they diminish. When they think it will make them more accepted to the audience, they inflate. Nowhere is this observed more clearly than in surveys about sexual behavior or alcohol/drugs/cigarette consumption. When I worked for five years tracking the performance of a sexual aid, we used to laugh at the responses to “Number of times per month” that couples were having sex. Straight, married males were having twice as much sex with their partners as straight, married females claimed. That doesn’t add up as it takes two to tango. Was it a case of “misremembering” or is it an issue of demure responses versus puffery? Then there was alcohol, a category I worked in for 7 years. It’s hard to believe this, but people over the age of 30 (historically) under-report the number of drinks they consume in a week. We received volume reports (by the ounce) for SKU sales in the US; it was consistently 20% higher than what self-reported consumption would have you believe. We found that when asked “On average how many alcoholic drinks do you consume in a week?” and faced with a scale of 0-14+, moving every respondent who did not say 0, up 1 notch, the responses were closer to actual market volumes. So what? What does this have to do with Trump?

Trump had a wicked stigma attached to him and the majority of reputable polls are conducted over the phone, person to person. When faced with choosing a candidate to a person you have never met, there are many who will say, “I don’t know” rather than admit they want to vote for a candidate that has social baggage. His numbers were underreported, not by polls but by respondents. The clue should have been in an undecided vote that carried as much as 10% with two weeks to play (all while the margin of error on the polls was 2-5% depending on methodology.) There is very little possibility that people were undecided on HRC, they knew who she was, they knew she was going to run, they knew what she stood for. They knew whether they could vote for her. The undecided vote was a collective of voters who were still unsure if they could pinch their nose and pull the lever for Trump or simply people unwilling to identify themselves as likely Trump voters. Everyone, including my friends working in Public Affairs research missed this. How?

5-10% undecided is not a large proportion heading into a presidential election. In fact versus 2008, it’s practically tiny. What was missed is that the numbers in 2008 and many previous campaigns without incumbents were among two candidates that had yet to hold major offices. This round missed the galvanizing component of 30 years of mud-slinging and vitriol attached to HRC (both for and against.)

It is therefore my opinion that the polls were inaccurate due to an unaccounted social behavior that rarely rears its head in candidate races. Social norm adjustment is common in polls of referenda that may include things like gambling, or other poorly perceived social issues. The candidates produced from a primary are rarely seen as socially unacceptable, but still electable. This was a perfect storm.

2. What did the media miss? How was this a surprise?

It’s rare that a coronation seems so assured only to see it be torn asunder by reality. In the information age, that simply shouldn’t happen. We have data, we have Nate Silver, and we have a 24 hour news cycle that rants ad nauseum to anyone who will tune in about expectations and swing states or worse, full blown predictions. What did they all miss?

There is a head-in-the-sand quotient to this all. The undercurrent of disrespect and the incredibly deep distaste for HRC in many suburbs, exurbs, and rural areas is not well understood by urbanites. I remember as a child in Michigan seeing bumper stickers in the mid-90s that read “Impeach Clinton and her husband too.” That underbelly was never addressed, or even given a hat tip during this campaign. As the DNC did everything in their power to amass a fractured movement behind a candidate, they forgot how divisive HRC was to Main St America. The love she saw in urban areas, and among young women clouded the picture of a truly divided populace, angry at what they saw was a system with Super Delegates cramming a bitter pill down their throat. They missed the Trump Democrats.

Trump Democrats seem almost impossible to anyone who saw the mandate that elected Obama. The wave of progressivism that ushered him into the White House twice was certainly strong enough to stand behind the first female major party candidate after it elected the first black president. The populace was woke, it was active and the progressive movement was strong. And yet, there were holes. True progressives wanted a true progressive, something we saw disappear when the wave of support behind Bernie was curtailed. There was a severe disenfranchisement that occurred when conspiracy theories gained traction, encouraged by more Podesta emails, that showed the unfair process that stiffed Bernie and produced Hillary. These were people not looking to stand behind Hillary, instead they were looking to alternatives, or worse yet looking to not vote at all. Meanwhile in the other camp, Trump was sweeping up the forgotten voters, the ones who felt their values were ignored for 8 years after being catered to under Bush. They felt disrespected, used, and worse than that, they were seeing things they cherished being diminished, teased or labeled as obtuse. When HRC called them deplorables, it galvanized them further to show her what they thought. And they did.

3. What does this meaning going forward? Are there lessons to be learned?

I believe there is always a lesson. Even when data is flat, there is something to be reasoned from it. Sometimes that reasoning is, “You haven’t done enough to produce actionable results.” Well the ‘good news’ in this is there is movement, there is a conclusion and there are action items that stem from this. Much like the campaigns built by Karl Rove we see a kicked hornet’s nest of previously disenfranchised voters, only instead of Evangelicals (who still played a part, but not nearly to the same levels as with Bush) being the impetus for a Trump victory, this time we saw it on illegal immigration. The media still does not want to separate illegal and legal immigration, for whatever reason, but clarity around this item would have shown that those “undecided” voters were seeing it as the top issue. Added to this is the inability for a candidate tied to a globalist economic agenda (as proven with the shock in the financial markets today) to connect with the activists that brought in the TEA party, and 99% movements. Put simply, a divisive, well-known candidate was unable to secure what she needed to triumph over her own reputation.

The only action item from all of this is that the primaries are simply the most important time to effect change and should not be taken lightly. This story plays out much like the narrative where a sports team needs to close a series out when they have a chance rather than let a pesky opponent stick around into a Game 7 when anything can happen. The best chance any of the US had at defeating Trump was in the RNC and DNC primary. That did not happen.

And for all of the talk that will happen today about a tight race and an active 3rd party and the electoral college, the best way to limit the influence of a third party is to provide a strong, electable candidate and the electoral college exists in the same way Congress exists, or even the way parliament exists in selecting a Prime Minister, its very nature is to reduce the influence of population skews that would have their own favor in mind over the favor of a sizeable other group. The House and Senate were composed of 2 officials, or a number of officials based on population so that states like Virginia and Connecticut could have fair voices in this union. There is a check and balance to the power of NYC, LA and Chicago that is provided with the Electoral College while still respecting the federated states that make up our nation.

/rant

Textured Soy Protein > PotbellyJoe and 42 others 11/09/2016 at 11:57

0

Hillary’s baggage helped turn rural/suburban white folks to Trump but Trump also did a good job of appealing to them.

Hillary squeaked out the popular vote in spite of all the demographic shifts going on that in theory should make it easier for a Democrat to win a national election, if we didn’t have the electoral college.

The big story here is that Hillary did a poor job of turning out the Democrat base. She still could’ve won in spite of the negative response from the rural/suburban white folks if she did a better job turning out minorities and not-strictly-Democrat liberals (i.e. Nader/Bernie enthusiasts).

If the Democrats are smart, they will get back on message for their platform. Only one candidate was saying things like “here is a problem out there in the world and I’m going to fix it,” and that was not Hillary. In spite of our increasingly negative politics, voters still want to hear that in order to turn out and vote.

PotbellyJoe and 42 others > Textured Soy Protein 11/09/2016 at 12:01

0

Sure, there’s part of that. I think the under-reporting polls led her to an air of hubris in states that turned into battlegrounds. 15 days ago, there is no way any nationally accepted poll had Trump taking OH and MI. Most didn’t even recognize that a battle was possible in Michigan. Why spend the money to get a base to the voting booth if the national polls show a comfortable margin?

RallyWrench > PotbellyJoe and 42 others 11/09/2016 at 12:04

0

“It’s hard to believe this, but people over the age of 30 (historically) underreport the number of drinks they consume in a week.”

Not sure if sarcasm. No one wants to admit they drink too much.

I’m now going to go back and read the rest of this, thanks for writing it.

PotbellyJoe and 42 others > RallyWrench 11/09/2016 at 12:06

0

Yes, full of sarcasm.

crowmolly > PotbellyJoe and 42 others 11/09/2016 at 12:08

0

Don’t know if I mentioned this on here, but years ago a company I worked with had some associations with an exit polling company. The polling place was very well respected although maybe not so well known. They had accurately predicted winners for many elections and had a sample that mirrored the US population breakdown almost exactly.

In 2004 they predicted a landslide win for John Kerry. There was shock and bewilderment when Bush won.

Reason? Same as you described. I guess you could kind of say it was a form of research bias.

For this election it appears that there were plenty of Trump supporters that were not vocal.

MUSASHI66 > PotbellyJoe and 42 others 11/09/2016 at 12:09

0

I agree wholeheartedly. Saying that you would vote for Trump around a liberal person was seen as a personal insult, as a sing of being dumb/racists/bigot/you fucking name it.

They way I see it, I am not happy that Trump won, I am happy that Hillary lost. Many people feel the same.

DNC shot themselves in the foot by not supporting Bernie. He had no baggage to speak off, and despite his way-too-socialist views, he would have done better IMO.

PotbellyJoe and 42 others > crowmolly 11/09/2016 at 12:10

1

If I remember correctly, 2-week polls in 2004 still had a very heavy undecided population as well, which again is something that should be lessened when there is a known candidate.

Textured Soy Protein > PotbellyJoe and 42 others 11/09/2016 at 12:11

2

You’re right, I’m in Wisconsin and the even the campaigns thought it would go solidly for Hillary. She didn’t visit here at all after the Dem primary and Trump cancelled a visit a few days ago because his campaign figured Hillary had it in the bag.

I think that hubris was part of what led to Hillary’s downfall in turning out the base. We’re on the same page.

Some anecdotal stuff, again from my state which very much followed what you’re saying about the polls:

Here in Madison which usually is of course extremely liberal, there was not the same enthusiasm for Hillary that we saw for Obama, Bernie, even Nader. People here will get up and shout from the rooftops about a liberal candidate they like and there were plenty of your “I’m with her” bumper stickers and whatnot, but it just wasn’t the same level of enthusiasm. People weren’t excited about Hillary.

I get to see a good chunk of the red parts of the state when I visit family, and let me tell you, there were bigass gigantic Trump and Paul Ryan signs everywhere. There was one house on the main drag in my mother-in-law’s large small town that had 3 different 10-15 foot Trump signs on the front lawn. I guess they’d rather look at the back of those signs than see traffic going by.

I see 3 things that happened:

1. Hillary turned off the whitefolks like you said.

2. Trump did a good job appealing to those same whitefolks.

3. Hillary didn’t do enough to turn out her own base to make up for it.

Krieger (@FSKrieger22) > PotbellyJoe and 42 others 11/09/2016 at 12:14

0

To Nate Silver’s credit, he did write this .

Well, Comey sure as hell ain’t gonna be indicted under the Hatch Act except perhaps as a final “fuck you” by the Obama admin.

And as to the folks cheering that the witch lost, well, enjoy your choice, you’ve earned it and I sincerely mean it from the bottom of my heart.

crowmolly > PotbellyJoe and 42 others 11/09/2016 at 12:16

0

I think any type of non-anonymous research (IDIs, TDIs, focus groups, even some quant stuff) might have this type of bias, depending on topic.

PotbellyJoe and 42 others > Krieger (@FSKrieger22) 11/09/2016 at 12:19

2

Yes, I wasn’t criticizing, or minimizing Nate Silver. The dude can run circles around me and my office full of underlings. What I meant in my statement is that under-reporting by respondents is nearly impossible to decipher and therefore means that even the greatest minds in statistical analysis can miss key items. Better follow-up research would have teased out the untruthitudes, but the authors of the polls would have to account for it and know what to ask. This is the value of hindsight.

PotbellyJoe and 42 others > crowmolly 11/09/2016 at 12:22

0

Sure, but in sex, drugs and rock and roll it’s better understood. I think when it comes to candidates vetted by a primary, it has traditionally been accepted that they are not socially stigmatized. I wonder if George Wallace under-performed in national polls versus actual turnout.

Axel-Ripper > PotbellyJoe and 42 others 11/09/2016 at 12:23

2

It seems to me like the Dems were so concerned about securing minority votes (Black, hispanic/latino, women) that they completely forgot about the white population, male and female. Especially the middle/lower-middle class ones, who were not only the ones who didn’t like her in the first place, but were EXACTLY the people Trump directly targeted. It was assumed she could pull all the women, because woman, but (pulling from coverage memory from last night) she was having a hard time securing even 50% of women voters.

Also, I blame the media a bit because they COMPLETELY IGNORED (and continue to ignore) this. The “Everyone who voted for Trump is a sexist, racist, misogynistic bigot” thought is exactly WHY Trump won. It is exactly like saying “everyone who supports BLM wants all cops dead.” If the dems (and Hamilton Nolan) want to figure out what went wrong, getting outside their urban echo chamber would help.

RallyWrench > PotbellyJoe and 42 others 11/09/2016 at 12:23

0

Carry on, then.

TheRealBicycleBuck > PotbellyJoe and 42 others 11/09/2016 at 12:29

0

“It’s hard to believe this, but people over the age of 30 (historically) under-report the number of drinks they consume in a week.”

One form of bias is the definition of a “drink.” I worked in alcohol research for a time. Not market research, but physical science. We studied the effects of alcohol on the body. One of the scientists was working on fetal alcohol syndrome (FAS). He was following up with one woman who claimed to consume only one drink per day, but her child had severe FAS. In the follow-up interview, he found the source of the problem. Her “one drink” per day was a 16-oz mason jar filled with moonshine. The standard definition of one drink is 1.5 oz of 40% ethanol, whether it is consumed as a single shot, one beer, or one glass of wine.

The researcher had to change his questionnaire to clearly define what he meant by “one drink.”

I wonder if the same definition error is causing the problem present in the market research.

PotbellyJoe and 42 others > TheRealBicycleBuck 11/09/2016 at 12:32

1

No, the surveys clearly define what is a drink. People under report because they don’t want people labelling them an alcoholic.

TheRealBicycleBuck > Textured Soy Protein 11/09/2016 at 12:36

1

My wife pointed out that Obama probably lost the hispanic (latino?) vote in Florida with his policies in Cuba. She’s not Cuban, but she is hispanic and she keeps tabs on the larger community.

I did some googling and found this to back her up.

DipodomysDeserti > PotbellyJoe and 42 others 11/09/2016 at 12:36

0

That’s the most well thought out, rationed rant I’ve ever read.

Textured Soy Protein > TheRealBicycleBuck 11/09/2016 at 12:45

0

That link is behind a paywall for me but I get what you’re saying. I think that’s one piece of the puzzle, but I keep coming back to the utter lack of “here’s what I’m going to do for you, dear voter” messaging from Hillary. I mean hell, I’m in white bread Wisconsin and I get Spanish-language commercials for the local gas & electric utility on my Pandora, but what was Hillary doing to appeal to minorities? I don’t even know.

Matt Nichelson > Axel-Ripper 11/09/2016 at 12:58

0

Deadspin is nothing but Gawker II now and that piece he wrote a little earlier from the Concourse is a prime example of it. Nolan isn’t the only one, though. Several of the staff have a seething hatred of what has gone on. 

TheRealBicycleBuck > Textured Soy Protein 11/09/2016 at 13:47

0

Ah, sorry about that. I forget about paywalls sometimes.

TheRealBicycleBuck > PotbellyJoe and 42 others 11/09/2016 at 13:49

0

Interesting. I’ve never seen one of the surveys.

PotbellyJoe and 42 others > TheRealBicycleBuck 11/09/2016 at 14:13

1

It’s typically something along the lines of:

“As you take this survey, you will be asked about drinking alcoholic beverages. For the purpose of this survey please define a drink as 5 oz of wine, 12 ounces of beer, or one shot or 1.5 ounces of liquor.”

etcetera etcetera.

The Lurktastic Opponaught > TheRealBicycleBuck 11/09/2016 at 15:56

2

You’re spot on. Being a scientist isn’t about answering questions. It’s about figuring out what question needs to be asked in the first place. Asking the wrong question may still net you a “right” answer, but it may not be the answer you need.

traitor joe > PotbellyJoe and 42 others 11/09/2016 at 16:18

0

Thanks for sharing your perspective. Well written and I enjoyed reading it.

In a Mini; let them mock me as My Mini Countryman is higher than you > Matt Nichelson 11/09/2016 at 16:53

1

The Jez authors and their commentariat are doing the definiton of the “liberal tears joke” right now. It’s not fun. They’re saying “fuck you” to 50% of the population (males) and are demonizing Trump’s supporters. Trump’s supporters feed off this, the more you attack, the more brazen they become (look at these past couple of months).

Matt Nichelson > In a Mini; let them mock me as My Mini Countryman is higher than you 11/09/2016 at 17:00

0

It’s indeed quite bad over there. Definitely a place to avoid for a while if that isn’t your thing.

gmporschenut also a fan of hondas > PotbellyJoe and 42 others 11/10/2016 at 21:56

0

!!! UNKNOWN CONTENT TYPE !!!

2008 should have been a clear indication of amount of support she received. In the mid 2000s I’d sometimes listen to an old right wing talk radio show that was completely stupid to pass the time for a good laugh. hrc was the “punchline” to half the statements. 20 years of ingraining that she is a walking satan is never going to get those voters.