Why You Should Ignore All the Polls

They Aren't Telling Anything You Need To Know

As of yesterday, the Economist’s assessment of the race. The numbers will probably shift a little today, but who cares.

Catch-up service:
In (Partial) Defence of Jeff Bezos
The Historian and the Culture Warrior
Giving Up On the Humanities Is a Mistake
The Orangutan Theory of Intelligence
Introducing ‘John & Paul’

Here’s the only information you need about the upcoming presidential election: it is extremely close and it could go either way.

That’s it.

Over the next few days, the noise generated by this election will reach its screeching, cacophonous climax. Here’s my instruction (to myself, as much as to you): every time you get an urge to consume information about it, go and do something else instead. Take a walk, do a headstand, read a poem, listen to Bach. I guarantee that you will feel - not just feel, be - better off in all the ways that matter.

If you’re betting large amounts of money on the outcome, or being paid to read about it, then you get a pass. Otherwise, no exemptions. My interdiction applies with special force to polling and voting data. Perhaps, like me, you’re tempted to gobble up every new poll of Pennsylvania, Michigan, Wisconsin; to track every tiny fluctuation in the national average; to dive into the cross-tabs for insights into the mindset of 18-24 year old Hispanic men in Arizona, to refresh your feeds for news about who’s voting early in North Carolina.

Perhaps you are weaving, out of the various threads of information you have gathered, compelling stories about what’s really going on - about how an unstoppable surge of pissed-off suburban women will deliver the key states to Harris, or working-class black men disillusioned with the Democrats will give Trump the edge - while citing granular demographic data from Montgomery or Wayne (the county in Michigan, not your friend Wayne, though maybe him too).

I get it. I’m vulnerable to all these temptations. But I have to tell you, dear reader, no matter how informed you have become, no matter how shrewd an interpreter of the evidence you are, no matter how deep your knowledge of the biases of individual pollsters or the different technical assumptions made by different models, it’s all a complete waste of your time. You don’t know the outcome and you can’t know it because there are too many variables, including possible polling errors in either direction. It’s very close and voters move in mysterious ways and that’s all we can say.

Believing otherwise is a kind of madness. I’ll explain what I mean by that in a minute - but first, do you know who secretly agrees with me?