House Forecasting Check-In: Just Two Weeks To Go!

A little more than two weeks 'til election day. Time for another check-in on the various House forecasting models.

As some may recall, the last time I did this was about a month ago. That was a few days before the Ford/Kavanaugh hearing and about two weeks before the Republicans hit their relative high point after that hearing.

The current reading of the forecasts is about two weeks after that high point and it is interesting to note that, at least in terms of House takeover chances and seat gains, the current forecasts have reverted to very close to where they were a month ago right before the Ford/Kavanaugh hearing took place. 538 and CBS Battleground are actually stronger on the Democrats' chances, while the Economist and Crosstab predictions have slightly weakened.

So, here are the current forecasts (readings from a month ago in parentheses, except for CNN which had not yet released a forecast at that time)

538:

probability Democrats take House: 85 percent (80)
predicted Democratic seat gain: 39 (37)
predicted Democratic popular vote margin: 8.9 (8.5)

Economist:

probability Democrats take House: 71 percent (71)
predicted Democratic seat gain: 28 (29)
predicted Democratic popular vote margin: 8.4 (8.6)

Crosstab/G. Elliot Morris:

probability Democrats take House: 75 percent (78)
predicted Democratic seat gain: 35 (38)
predicted Democratic popular vote margin: 8.9 (9.2)

CBS Battleground:

probability Democrats take House: no estimate
predicted Democratic seat gain: 31 (29)
predicted Democratic popular vote margin: no prediction

CNN

probability Democrats take House: no estimate
predicted Democratic seat gain: 31
predicted Democratic popular vote margin: no prediction

Given these data, what are we to make the of the spate of stories downgrading Democrats' chances for a "blue wave"? It depends on the story but one of the most common points made is that Democrats' chances to take the Senate have eroded. That appears to be true but of course that was never very probable anyway; only the very highest of blue waves could possibly have got that done.

The most interesting point made is that Democrats' chances of really big gains in the House may have eroded. That is, even if the Democrats' chances of gaining enough seats to take the House aren't much changed, their chances of gaining, say, 40-60 seats have dropped.

Those who make this assessment base their view on the perception that more reddish, rural districts have had their Republican bases engaged by the Kavanaugh fight and Trump's grandstanding--as well as perhaps the sheer proximity of the election--and are now much less susceptible to Democratic insurgents.

That could be true though there are some countervailing factors that push the House calculus in the opposite direction. These include the Democrats' massive fundraising advantage--identified by Nate Silver as a factor that could result in greater Democratic than expected--and the poor performance of GOP Senate and governor candidates in the Midwest which could hurt downballot Congressional candidates.

Perhaps this is why the 538 model still gives the Democrats as much chance of exceeding a 61 seat gain (10 percent) as dropping below a 19 seat gain.