Wanna Know How Close This Is?
So, we’re 21 days from election day. Since the first of the month, the “momentum” indicators, which for months have clearly benefitted Democrats, started to flip. Democrats still seem ready to crawl over glass to vote; but for the first time, the Republicans seem ready to do the same. How this plays out in the House and the Senate seats is now moving out of the ability of the election prognosticators to predict, and into the hands of the on-the-ground pollsters. (I think of this a bit like a manned rocket launch, where at some point during the countdown, Launch Control has to “hand over” control of the vehicle to the astronauts who are actually flying the thing.)
The Senate, of course, is easier to predict for a couple of reasons. First, larger statewide populations yield better polling results with smaller margins of error. Secondly, because there are fewer of them, the major, professional polling firms do the polls.
House districts are a different bird. First of all, smaller populations mean less certainty and more volatility; if the 200-odd Jones clan decides that they’re having their family reunion on election day, that can skew a result. So can particularly bad weather in a particular district. These effects are muted somewhat by early voting, but they still exist.
According to the RCP aggregation, if the election were held today, the GOP would pick up three seats, which is welcome news for them, because the 2020 and 2022 maps are not GOP friendly; they’ll need a buffer count to hold control through to 2024, when the map turns GOP friendly again. The tossup seats look like this:
Arizona: The RCP aggregate shows McSally with a minuscule .3% lead. The most recent poll, taken by the ABC local affiliate in conjunction with a professional firm, shows McSally (who is such a good candidate compared to Sinema that this shouldn’t even be close) moving outside of the margin of error with a 6 pt lead. There is currently a negative news dustup around Sinema that should lead to McSally consolidating that lead. (Sinema is a former Green Party nutcase who has been backpedaling for some years to make herself more electable; however, the Internet is forever.)
Florida: The RCP aggregation shows a dead heat. The most recent poll shows Rick Scott with a 2 point lead; the caveat here is that that was a local poll. No telling what happens on this one.
Indiana: Donnelly leads all polls taken, but all are within the margin of error. Again, no telling. Donnelly seems like a nice guy, and voters seem to respond to him, even in an R leaning state.
Missouri. Hawley has a miniscule .5% lead over McCaskill; the latest poll shows them tied. Still another one that can’t be divined. One can’t help but pull for Hawley here; McCaskill is truly an abominable swamp creature that would have been put out to pasture six years ago, had the Tea Party not screwed up the race by helping nominate a nutcase on the GOP side. There is dustup in the McCaskill campaign right now from Project Veritas that does not yet appear in any polls.
Montana. Tester has a 3% lead in the RCP aggregation, and like Donnelly, has held that small advantage, within the error margin for some time.
Nevada. Dean Heller shows a 1.7% margin on the RCP rollup, but like Arizona, seems to be moving outside of the margin of error in recent polls.
And if you’re wondering, Heitkamp’s North Dakota seat, Ted Cruz, and Marsha Blackburn have all moved into “leans GOP”. That all leads to a +3 seat result for the GOP in the Senate.
Now, the hard part. The House.
The GOP have 235 seats in the House. 218 is required for a majority. So, the GOP can lose 17 seats and still hold the House.
The current “board” at RCP shows 205 seats being likely/lean/safe Democrat, and 199 seats being likely/lean/safe GOP, with 30 seats in tossup. A little simple math tells you that if the Dems and GOP split the tossups down the middle, that leads to a 220–215 Democrat House.
The good news is that a Dem house with a five seat margin is not enough to lose sleep over. Some of the right wing pundits are painting a nuclear scenario where the Dems will start impeaching everything in sight if they win the House; not going to happen. Pelosi is not stupid, and she knows that if she has the House with that small of a margin, then her majority depends on holding on to Democrat seats in deep-red districts. So, she’ll avoid having those members make controversial votes.
Putting that another way, she won’t take an impeachment vote on Trump or Kavanaugh if she has to choose between whipping a dozen red-district Democrats to take a vote that will cost them that seat in 2020, or lose the vote. Her posture will be “investigate, but not impeach.” Take it to the bank.
So, to the topic of the tossup districts. What will they do? The answer is “hard to tell”. The following is a drilldown on the polling in each of those districts (all of which, save one, are currently held by the GOP). Keep in mind that district polling is notoriously variable for the reasons stated earlier, and that this election will have surprises for us not just from the tossup districts, but from a few lean GOP and lean Dem districts who suddenly don’t lean that way on Election Night.
Here are the tossups and their current leanings:
So, add THAT up. The GOP has 199 lean/likely/safe seats. If all those leaners hold (R), that’s 217 seats. The Dems, OTOH, have 205 lean/likely/safe seats. If all their leaners hold (D), THEY will have 217 seats.
So, hello, Virginia 5, which has no polls currently listed in RCP. Is it possible that this election is so close for control of the House that it comes down to a single district for which no professional (or otherwise) polling exists? (I’m sure it exists, somewhere, but not on RCP, it should be said.)
Virginia 5 is a central Virginia district where both candidates have never held elected office before. Denver Riggleman (R) is a businessman (distillery owner) and a former Air Force intel officer. Leslie Cockburn (D) is an Emmy-winning journalist and the mother of “House” actress Olivia Wilde. Both are Boomers.
So, we’re left with only being able to analyze the district itself, which Five Thirty Eight does here:
Virginia 5th - 2018 House Forecast
Adjusted polls R+2.6 CANTOR R+0.8 Fundamentals R+7.4 Experts R+2.6 R+1.0 R+3.1 <0.1 <0.1 Historical adjustment #Similar…
….and gives the (R) a 66% chance of winning.
So, when you eliminate all the leaners, and assume that anyone with a polling lead is going to win that race, AND accept the prognosticators view of VA-5…….that all leads to a 218–217 Republican house.
Close enough for ya? :-)
Keep in mind there’s a better chance that you win the Powerball jackpot than that all of the above works out like that. :-)