Polling error and bias are not correlated across election cycles.

in polling •  8 months ago 

image.png

I'd never pretend to know more about politics than Dave Wasserman, but this is incorrect or misleading regarding polling. Polling error and bias are not correlated across election cycles. It is tempting to do so, but the data (albeit a small sample) does not bear it out.

Biden is down against Trump currently, but not by a large amount. Times/Siena was quite accurate in the 2022 Midterms, but it doesn't necessarily mean they will be again in 2024. We just don't know. At best we can aggregate the polling across firms to get a reasonable guess. With that current aggregate, Biden is down, but it wouldn't take an absurd polling error or change to favor Biden.

Of course the other wrinkle is that with the electoral college system, states are what matter. And the way voters are distributed, Republicans tend to have an advantage such that just beating in the polls is not enough.

I wouldn't want to be Biden right now, but I don't think anyone should have certainty about the state of the election this far out. And it isn't insurmountable for Biden.

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!