Political polls aren't reality. They are attempts to influence reality.

Honestly, ask yourself: Why would they be accurate? What would be the point?

Political polls are nonsense, right? I mean, nobody is supposed to actually believe their results, right? Do you?

All polls are highly suspect, in my opinion. About anything. First of all, my Austrian-economics-influenced mind doesn’t trust any claims of preference that aren’t actually demonstrated in the world. I’m not interested in what you “say” you do (or will do). I’m not convinced until you actually do it. Show me.

And then there’s the methodology problem. Are these pollsters actually doing the work of getting an accurate cross-section of the public along all the possible dimensions that would be required to get an accurate sampling of the population?

How would they even do that? How do they contact these people? How do they accurately assess the demographic/ideological information required? Do they still rely on land-line phones? How do they account for the large segment of the population that doesn’t talk to pollsters? How do they handle people who like to lie to pollsters, just for fun?

I have serious theoretical and methodological questions.

But leaving those quibbles aside, I can imagine that polling (if done carefully) might be useful in commercial market research to test consumer preferences for various things. It could be worth taking a shot at it.

Apparently a lot of big companies do so, so they must at least think they see a benefit. And in this use case, the polling company is incentivized to be accurate. Otherwise, the company won’t use them in the future.

But political polls? Why would they be accurate? Why would the polling company even want them to be accurate? First of all, who’s paying for political polls?

According to the “AI” built into my web browser (and this CNBC article), political polls are usually loss leaders for polling companies. They are used to promote their own brand in order to attract commercial polling business, or to sell their consulting services to campaigns.

So, if that’s the case, why bother being accurate, especially early in the campaign cycle? Why not take advantage of this period where it doesn’t really matter if you’re accurate or not?

After all, every one of these polling companies is staffed, owned, and managed by opinionated human beings, so of course each company has certain biases.

Or do you think a Fox News poll or an MSNBC poll about the presidential race are going to be completely unbiased? Do you think they’ll get the same results?

I doubt it.

I think early on in an election cycle, political polls will be heavily biased in favor of the candidate preferred by the polling entity (or perhaps their biggest corporate clients). This can have the effect of making certain candidates seem more credible to voters who are easily “herded”.

We saw this quite powerfully with Kamala Harris, both in her brief 2020 presidential primary bid and after the “soft coup” that saw her swapped in for Joe Biden after it was too late for Democratic primary voters to do anything about it.

In both instances, the corporate media (and whoever pulls their strings) showed Harris with surprising support in “the polls”. But this was all manufactured, hot air, “AstroTurfed” nonsense. Very early in the 2020 primaries, real voters demonstrated that this support just… didn’t exist.

And right now, it looks like the same thing is happening. After being touted as “joy incarnate” and insanely popular in the polls, now that election day is less than three weeks away, those poll numbers are starting to move in a more realistic direction.

This happens a lot with political polls. After all, if these are loss leaders designed to promote the brand name of the polling company, they can’t be too out of whack with the actual election results in the weeks leading up to the vote.

They can be sloppy and biased when the election is months away. But weeks or days out, they need to quit trying to manipulate the public, and actually make an effort to gauge the public.

Doubt my analysis? Here’s a little question that might shift your intuitions in my direction. If these political polls are accurate, why do campaigns do internal polling? Why go through the hassle and expense? Why not just rely on all these very very many polls that are out there for free?

They can’t. They know those polls are nonsense.

And this morning I’m feeling especially cynical, so I’ll share another nagging concern I have about political polls. I believe that the Deep State has very strong influence in the corporate media. I believe this influence extends to most of these polling companies.

If the Deep State has managed to corrupt and (at least partially) control our election systems in key battleground states enough to manipulate the results, then they have a huge incentive in making sure that by election day, the polls show the race close enough for their preferred result to be believable.

They need the polls to not only be in the “margin of error”, but the “margin of cheating”.

This concerns me quite a bit. And now, perhaps, it concerns you. Sorry about that.

Naturally,

Adam

PS: Oh, and if you want a better understanding of why people have such seemingly intractable deep biases, I can’t recommend this book by Jonathan Haidt enough. It’s called The Righteous Mind. Check it out.