It has been discouraging this past week to realize that shouting heads in the Democratic Party and the punditry class really don’t understand elections. Their arguments are based on a very wrong idea — that a decision to replace Biden is like hiring a new CEO. Even the ordinarily wiser Nate Silver has failed in this regard.

They also seem clueless that even in companies, these managerial hiring processes fail far more often than they work. Corporate histories are littered with CEOs who were the logical and wrong choice.

Democratic pundits and shouting heads have not shown they know that elections are not logical, rational processes — they are not managerial. This leads us to complexity. Elections take place in a highly complex world. For nearly 70 years, though, they have been somewhat stable as candidates on both sides acted on the same “logical and reasonable” mythologies in most cases. But the lessons learned from those years can fail – and they seem to be failing because Trump is neither logical nor rational — he’s unstable and unreasonable.

That means today’s elections take place amid high instability. Further, amid complexity like this, sometimes only one change will causes previously reliable assumptions to end in disaster — assumptions like those used in polling as well as aggregated studies of poll results.

Irrationality in Polling

I like to read the survey questionnaires with polls to see how the questions were posed. After all, survey results always depend on how questions are posed and the order in which they are posed. This year, one particular approach is surprising and I’ve found it in every poll I’ve investigated:  They start by asking participants about their feelings about whether the country is on the right path OR to rate President Biden’s job performance.

This is probably a polling “best practice” developed in the modern period of presidential polling from around 1950. As a best practice, it has been adopted on logic like “history shows us this effectively judges likely voting in the future.” Notice, also, the questions assume voter choice is a primarily logical/rational activity.

Those assumptions are no longer valid:

• No presidential polls have ever been tested with an incumbent opposed by a previous incumbent. The last time the US even saw such a situation was when Teddy Roosevelt ran in 1912 — well before modern polling.
• The most impassioned Trump voters are NOT making their choices based on logic — but on the sense of emotional connection and satisfaction “putting it to the man” which he leverages.
• Polling best practices were built before structural changes radically interfered with polling practice. Today, all polls have to be adjusted because the sample of “the people a pollster can get in touch with” is not a valid sample of “all likely voters.” Why? Cell phones, internet, no land lines, general refusal to participate in research etc… Each poll tells us they wisely make these adjustments. I am skeptical as it’s only become harder.

The difficulty with these past practices was evident somewhat in 2016 and more evident in 2020. In neither case, though, did it lead polls to be entirely off-base.

The specific conditions this year are dramatically different. For the first time since polling began, the incumbent president faces an incumbent former president — one who is also irrational.

This makes the inherent design of the polls hurt Biden numbers while helping Trump numbers. After all, voters know a LOT about Trump as well as Biden. But they are only seriously asked about Biden. Today, NO poll should open with a Biden job approval question because it prepares the participant to be more negative about Biden than about Trump. These polls no longer accurately measure voter intent because the questions manipulate voter intent before asking “who will you vote for.”

How bad is the problem? I have no idea. In my work for companies betting millions on new product introductions, an error like this would lead me to throw out all answers to “who are you likely to vote for.” Despite this there is probably some validity in the numbers but they are also highly invalid. How much might they overstate Trump support by? We don’t know. 2 points? 5 Points? 10 points? We don’t know.

Wise pollsters would be investigating this challenge using exploratory polls with a variety of designs. I don’t know if this is happening but I suspect it isn’t — these investigations are expensive and the pollsters are making nice money while their sponsoring media outlets are also making nice money.

Note, also, that these errors defy statistical games like those loved by renowned Bayesian Nate Silver or by NN Taleb. We know the errors in which case statistics and probabilities won’t discover the truth — errors embedded in the vast majority of respected polls. (I confirmed this once more with the new NY Times poll out day — July 4th.)

Shifting to Campaigning: Voter Choice is Illogical

The majority of political professionals and pundits seem to believe voting takes place in a heavily logical and rational world. Of course, within the halls of government maybe this makes sense. With the general US population it doesn’t.

James Carville knew this in Bill Clinton’s first race with his focus on “it’s the economy, stupid.” He knew where connections lie. And he leveraged an emotional win with George HW Bush’s error not knowing the price of a loaf of bread.

Today, the surprising support for a convicted criminal, known rapist, and congenital liar defies logic. That support, though, seems partly due to the great many people in the US who are fed up with the inhumanness of theories like “life is logical and rational.” It isn’t. Let’s get over it.

Here is the most important truth about emotions and the electorate today:

Democrats cannot replace Joe Biden and still win — no matter who they choose.

Yes, we see polls which pose hypotheticals like “what if Trump was running against ___” (pick your favorite alternate). Companies know these answers are meaningless and, the smart ones, never attempt to predict future sales from questions like “will you someday buy this product?” Companies who choose to trust these answers always end in defeat.

These are all untested products the voters barely know. Their names are, mostly, only known to the Democratic elite and no voter has seen ANY of the alternatives on the campaign trail OR facing Trump. Unless someone has a surprising magical ability, the polls on the issue are entirely wrong — 100% wrong. They also don’t tell us about candidate X vs Biden. This approach, in fact, is prejudiced AGAINST Biden.

Emotions Will Win

All these shouting heads assume elections are won by logic and reason but they aren’t as we saw with Trump in 2016. That means Biden and the Democrats will only win based on emotions. What emotions will work? Let’s start with those that won’t work.

• Protecting the constitution: Trump has muddled this issue too much.
• Women’s rights and Abortion: Quite helpful. But it’s not a fresh and new message.
• Arguing over exact economic results: Best way to make voters yawn.
• No, Really. Joe is Perfectly Healthy: There is not a “test” which gets past concerns. Biden needs to campaign effectively — using his underdog status now as the foundation for impressing the American people. Really, people love to see an underdog overcome nearly insurmountable odds.
• Look, Joe Rides a Bicycle! This has been quite useful. But the past 7 days of shouting head attacks made it a moot point.

As an advertising guy, let me suggest the one emotion which I believe would work:

Trump is a Coward

Americans do not like cowards. They do not like people who evade responsibility — who refuse to own up to their errors. Trump has exceptional abilities squirming minimal success out of disasters. He lacks the courage, though, to identify the successes which matter and ensure we reach them. The US needs a courageous leader today.

Making this approach better, the US doesn’t need the peak of physical fitness. FDR is exhibit A here. JFK was even severly hindered by back problems he worked hard to hide from the American people. All I’ll say is most marathoner’s I’ve met aren’t capable of being president — they simply can’t do it.

Courage is an important part of fitness for office — one where Biden wins and Trump ignominiously loses. Only leaders who are courageous are fit for office — they must thrive despite the vast range of unusual circumstances the world might throw at them.

Biden is courageous. Trump isn’t. As an advertising guy, I’d suggest Biden’s campaign look for an equivalent of LBJ’s atom bomb ad which drove emotions in his 1964 win over Barry Goldwater.

Why am I writing? Will this post make any difference? Who knows. I had to get this off my chest and hope someone out there will listen.

Have a Happy 4th and be well.

©2024 Doug Garnett