The number of polls showing a perfect tie in the presidential election is incredibly high.
It doesn’t mean “there are a lot”, it literally means “there are an incredible amount”.
Polling’s track record these days is about as reliable as a coin toss. They completely discounted President Trump’s 2016 victory. The results in 2020 were even worse, with Biden expected to win in a landslide. They promised us a “red wave” in 2022, but it turned out to be a bigger ripple. And let’s not forget that they completely missed Brexit across the pond.
What interested me was that there was a pattern to these failures. Polls aren’t just wrong; In a world without polls, public opinion is wrong, just as you would expect if you followed the conventional wisdom of the moment.
and people are political
Recall the example above that started in 2016. The media consensus was clear. Trump had zero chances. What about public opinion polls? Surprise, surprise – they showed just that. In 2020, after four years of media pressure and disruption caused by the coronavirus, polls showed Trump winning by a landslide. In Britain, the educated elite could not imagine that their own citizens would actually vote to leave the EU. Again, the polls agreed.
Pollsters are quick to blame technical flaws for failures. “Shy Trump supporters” won’t answer the phone. They overcounted college-educated voters. Voting patterns have changed. But maybe there’s a simpler explanation. They are human beings who are exposed to the same prejudices as us.
The real poll question isn’t about math. It’s about human nature.
Today, conventional wisdom says this race is too close to impossible to run. Given the standard sampling error of polls, even if the race was actually an exact 50-50 tie, the poll would be wide-ranging; average Approximately 3% difference. That’s not what we’re seeing at all, just a tight clustering of polls with almost half of them showing exact ties as of today.
RELATED: White House reportedly altered official transcript of Biden’s ‘garbage’ comments
There’s a term in the polling industry: “herding” to describe when surveys mysteriously cluster around the same numbers. And it’s surprising when pollsters reaffirm their methodology after seeing results that differ from other pollsters. – Find a reason to align towards consensus.
Polling analyst Nate Silver, who has spent his career primarily doing numbers-based polling, is a little flustered when he notices a clear trend. “I don’t really trust pollsters,” he said on the podcast. “When you sample 800 people over dozens of surveys, not every number comes with an exact 1 point lead. You’re lying! You’re putting your finger on the scale!”
He’s right about grazing. Pollsters are mortally afraid of appearing foolish on election night, and they can avoid that by keeping their numbers close to everyone else’s. The analogy of running safely through a herd of animals is spot on.
How it actually works
But a whole swarm of pollsters everytime There is a finger on the scale. There is no such thing as raw data.
Opinion polls involve more than just counting responses; they involve hundreds of decisions. How many young voters will participate? What percentage of the electorate will be college-educated women? Should we base our considerations on past voting behavior?
These are not clear-cut mathematical decisions. They are intuitions, educated guesses about human behavior. And like all intuitions, they are influenced by what we believe to be true.
RELATED: About 63 million voters have already cast their votes
That’s just human nature. We all tend to see what we expect and find ways to justify our existing beliefs. Pollsters, despite their scientific pretensions, are not immune to these psychological functions.
Such bias creeps in when you have to make dozens of decisions when designing and interpreting polls. If you “know”, consciously or not, that Trump can’t win, you will choose methodologies that support that belief. If you are “confident” that the race is close, then you “refine” your hypothesis until it shows exactly that.
I’m going to go out on a limb here and say the whole herd is wrong. This is just a hunch as the data clearly doesn’t match up, but I don’t think this is a close race. I think the trends of 2016 and 2022 will continue and that we are vastly underestimating Trump’s power. Of course, you can’t say that out loud at most Washington insider cocktail parties.
So when you see another poll showing an exact tie in the presidential election, remember that behind those decimal places and margin of error calculations are people making decisions. And they, like you and me, can’t help but be influenced by what they think they already know.
