Nov.04

electoral-vote.com

Dem 50

Ties 2

GOP 48

But Wait, There's More!

There is a definite possibility (see above) that Trumpy voters are being undersampled, that the Republicans will have a better night on Tuesday than predicted (again), and that the pollsters will end up with egg on their faces (again).

On the other hand, there is also a definite possibility that things may swing in the other direction. In the special elections held this year, Democrats have being doing surprisingly well and that has nothing to do with Republicans slamming down the phone. Of course, there have been only a handful of them, so they don't give us a lot of data to work with.

There are a bunch of other factors that could introduce significant uncertainty into the polls. Some of the biggies:

• When it comes to the Trumpers, it is possible that some houses will overcorrect to avoid yet another embarrassment. Emerson leaps to mind.
  
  
• Nobody really knows what the impact of the Dobbs decision will be. If the electorate was, say, 54% Democratic in 2018, then most pollster models are going to base their projections on the notion that it will be similar in 2022. And "how likely are you to vote" questions can only move the projection a little bit. But if Democratic turnout is way up (as it has been in the special elections), then the models will be off and Democratic strength will be underestimated.
  
  
• Young voters tend to be fickle when it comes to midterms. There is at least some evidence that, this year, they will be less so due to abortion and student-loan forgiveness. If the young whippersnappers show up, that will help the blue team a lot.
  
  
• The country looks to be in the midst of a partisan shift, with blue-collar workers headed to the red team and educated suburbanites moving to the blue team. The latter tend to be more reliable voters in midterm elections than the former. It may soon be the case that the convention wisdom shifts from "Democrats don't show up for midterms" to "Republicans don't show up for midterms."

The upshot is that the polls are so close, and there are so many X-factors, that the crystal ball is very murky right now. Only a red tsunami, or a blue tsunami, would be a real surprise. Anything else is well within the realm of possibility.

That said, there is one other possibility for projecting the elections: The Wisdom of Crowds. As James Surowiecki observed, large groups tend to be better at making estimates than the individual members are. The general idea is that, in the aggregate, various errors tend to cancel out, leading to a purer and more accurate response.

Surowiecki's lead example, in his book, is people guessing the weight of a cow. However, there's no reason the same basic precept shouldn't apply to election projections. So, as you might expect, we put together a survey for readers to make their best guesses as to how the House and the Senate will end up. The survey will be open until 10:00 p.m. PT on Monday; we'll share the results on Tuesday. (Z)