Three advance indicators of the House outcome in 2018

July 11, 2018 by Sam Wang

Three indicators point, with varying degrees of strength, toward a Democratic takeover of the House in this November’s election. Two of them are the usual suspects: the generic Congressional ballot and special elections. There’s a third one that Harry Enten of CNN has noticed: which side is *not* talking about its internal polls.

I’m still processing this and other information. For a summary of my current thoughts, read this thread.

If you have other data-based indicators, please share them in comments. I welcome data-based analysis in the other direction.

Topics:

3 Comments

Pechmerle says:

Speaking of activism (note smooth segue), any thoughts on which Senate races to donate in to have best impact on the outcome?

Sam Wang says:

Yes, this is very important! I should post something soon. Current thoughts follow.
I see six knife-edge Senate races: Nevada (D-Rosen v. R-Heller), North Dakota (D-Heitkamp v. R-Cramer), Missouri (D-McCaskill v. R-Hawley), Florida (D-Nelson v. R-Scott), Indiana (D-Donnelly v. R-Braun), and Tennessee (D-Bredesen v. R-Blackburn).
Not counting these races, Democrats are favored to end up with 46 seats, Republicans with 48. To take control, Democrats would have to win five out of the six knife-edge races. This is a tall order but not at all impossible. Races that look close at this point in the season tend to end up breaking in the same direction.
In addition, one might keep an eye on the Arizona and New Jersey races, though I think Democrats will likely win both.
For further reading, see David Byler: https://www.weeklystandard.com/swingseat-senate-election-model/. For an interactive chart, see 270toWin: https://www.270towin.com/2018-senate-election/

LondonYoung says:

In ye olde days people looked at “Direction of Country”.
“Wrong track > 60” = bad for party in the White House
It is usually pretty correlated to generic ballot and prez approval, but it is leaning more GOP this cycle.

Leave a Reply

Your email address will not be published. Required fields are marked *