The Optimistic Leftist: It's Election Forecasting Time!

Sunday, August 19, 2018

It's Election Forecasting Time!

Election forecasting season is heating up with the release of 538's spiffy new House forecasting model. For those who have not yet seen it, their standard model (they have two alternate versions) gives the Democrats a 3 in 4 chance (75.3 percent) of taking the House. The average Democratic gain is projected to be 35 seats. As a nice bonus you can look up the chances that Democrats will take any particular seat both through maps and lists.

While the 538 forecast is the new and shiny, there are several other credible models that get much the same results with less complicated methodologies. The Economist model, which has been running since late spring, gives the Democrats a 70 percent chance of taking the House. They project an average Democratic gain of 29 seats.

G. Elliott Morris' Crosstab site has also been running a model for quite awhile. He gives Democrats a 76 percent chance of taking the House (no specific seat gain projected).

So everybody seems to singing from the same hymnal which is reassuring. It hardly needs emphasizing that these models generate probabilities not certainties and that the improbable sometimes does happen. But the agreement among models and the fairly high probabilities assigned to Democratic takeover simply reflect the fact that almost all of the data we have right now is telling a story favorable to the Democrats.

Just how favorable the story is was emphasized in some interesting remarks by Cook Political Report's David Wasserman--as astute and careful an analyst of House elections as you can find--in an interview with Jonathan Swan of Axios:

"Dave Wasserman, the Cook Political Report's House analyst, says the most under-covered aspect of 2018 is that "a blue wave is obscuring a red exodus." Republican House members are retiring at a startling clip — a trend that senior White House adviser Kellyanne Conway told me earlier this year was worrying her more than any other trend affecting the midterms.

There are 43 Republican seats now without an incumbent on the ballot. That's more than one out of every six Republicans in the House — a record in at least a century, Wasserman says.

Just in the past eight months, the number of vulnerable Republican seats has almost doubled, according to Wasserman. Democrats need to win 23 seats to claim control of the House. Today, the Cook Political Report rates 37 Republican-held seats as toss-ups or worse. At the beginning of the year, it was only 20.

Wasserman says the most important sign that 2018 will be a "wave" year — with Democrats winning control of the House — is the intensity gap between the two parties. In polls, Democrats consistently rate their interest in voting as significantly higher than Republicans. And Democrats have voted in extraordinary numbers in the special elections held the past year, despite Republicans holding on to win almost all of these races.

"There's a bit of over-caution, perhaps, on the part of the punditocracy, after what happened in 2016," Wasserman told Axios. "But if anything most media could be under-rating Democrats' potential to gain a lot of seats. They could be caught being cautious in the wrong direction."

So it looks pretty good. But it ain't over 'til it's over.