What we learned from the Tennesse House special election

The counting is basically done in Tennessee and Republican Van Epps will win by 9 points with the margin at 53.9% to 45%.

It was a comfortable win but not nearly as comfortable as the +21 spread in last year’s general election. So that’s good news for Democrats but I don’t think that’s a takeaway tonight.

Ultimately, this vote ended up about where prediction markets saw it and was maybe a tad better than feared for Republicans.

What this shows is that it was a pretty good night for betting markets.

The loser on the night? The New York Times election model.

That’s my real takeaway from this. If you were watching it live, it had Van Epps at just +3 points with around 35% of the vote counted. Now this model is supposed to be done on district-levels, so should account for where the votes are coming from.

With 35% of the vote in, that’s a huge sample and a good model should have been able to predict the ultimate result to 1-2 percentage points. Instead, the NYT model was off by 6 full points. I just don’t think that kind of model is going to be very useful in next year’s midterms if you’re thinking about trading them live.

Beyond that, the district level numbers show a clear shift toward Democrats and some pretty strong enthusiasm in Democratic-leaning districts.

Again, I don’t see this as anything market moving as it was already assumed given Trump’s low approval ratings. We are also a long way away from the midterms. On its own, I don’t think this is going to make Republicans push harder on Trump but I would be watching to see if Johnson faces pressure as House Speaker.

This article was written by Adam Button at investinglive.com.