Discussion about this post

User's avatar
Will, from Cal's avatar

Yeah... being a numbers nerd, I was looking through the cross-tabs, and... there was really no reason to publish that poll other than they already spent money on it. The details are pretty unbelievable, and not because I don't want to believe them, but because they are unbelievable. They didn't actually do it state-by-state, the margin of error is huge, the likely voter screen is way off, the swings predicted would be way bigger than anything in recent history, they don't match up with other recent polls or midterm/special results, and - this is key - there are 5% more self-reported Rep-leaning identifiers polled and more than twice as many self-identified conservatives.

They are an excellent outfit, but everyone gets unreliable results once in a while, and - pardon the language, but it pisses me off royally that because NYT has such a readership it is going to freak out a lot of people who are not aware just how far off these results are from likely actual reality.

Expand full comment
maren showkeir's avatar

Thank you SO MUCH for giving us the Mike Podhorzer article that says that political polls are completely useless. They are.

I am a long-time print journalist, and I have been screaming into the void about the other point he makest: The polls are not useful, and are even dangerous. And it doesn't serve citizens to cover political races like horse races. The stakes are too high.

Maybe we should turn some of our attention/energies into "lobbying" media organizations to be more responsible in their reporting. If we unleashed a raft of letters/calls/emails to news organizations pointing out how harmful and not useful this coverage is, maybe editors would think more carefully about future coverage.

I plan to send an email right to the NYT right now.

Thanks for all you do.

Expand full comment
19 more comments...

No posts