Good Friday Morning! There are so many things I could write about this week — birthright citizenship, and the 14th Amendment may still get a write-up later on — but the looming story is the midterms. And I’m going to lay out where things stand now, one week out, and what my predictions are for the House. I’ll only make individual race predictions in the Senate. There are too many House races to call. Links to follow.
Where you can find me this week
Make sure to sign up for the Conservative Institute’s daily newsletter. You can also go to their Facebook page. You can join Ricochet here. And I do recommend their ever-growing network of podcasts, which you can find on all popular podcast platforms. They have a show for every topic you can imagine, and the list continues to grow.
My piece in the aftermath of the pipe bomb story. I walk through some examples of political violence in American history and what it tells us about our current age.
A case in the EU where they upheld a blasphemy law against a woman.
We’re one week away from the midterm elections, and as such, it’s prediction time.
Conventional wisdom, no matter what website or model you use, says that Democrats will win control of the House and Republicans control the Senate. The quibbling is over how big or small that power ends up being. Most conventional wisdom I’ve seen has Democrats winning between 30-40 House seats, and Republicans gaining 2-4 Senate seats.
To get how conventional wisdom gets here, I think it’s best to go back over the full points I laid out last week, and update the numbers as needed:
- Midterms are usually a referendum on the governing party, and the GOP holds all three branches of government
- Midterms are bad for the party in power. The party in power typically loses around 25 seats in a midterm election in the House.
- Donald Trump, though an outlier in many respects, is still a President for the governing party
- Trump has a below-water approval rating. FiveThirtyEight has him at about 42% approval, and RealClearPolitics has him close to 44%.
- The generic ballot is vital in midterms because it tells you trendlines towards a given party. Due to population and demographics, Democrats typically hold a 6 point lead in the generic ballot. Right now, FiveThirtyEight gives Democrats an 8 point advantage in the generic ballot, and RealClearPolitics has Democrats with a 7.5 point lead.
- Anecdotal evidence heading into the midterms tends to benefit Democrats: enthusiasm, fundraising numbers, and the number of opportunities on the map all point the Democrats direction. Republicans currently have 46 races described as a toss-up or worse for them, while Democrats only have 3 in that category according to the Cook Political Report.
- Finally, I check the models. Nate Silver is the gold standard here, and he currently has Republicans at 85%, or 5 in 6 chances to hold the Senate, with only a 15% chance, or 1 in 6 chance at retaining the House. Decision Desk HQ, a new entry in modeling, has the GOP at 5% odds to keep the House and 91% odds to keep the Senate.
As an aside, I’ll be following Decision Desk HQ, both on their website and their live feed on Buzzfeed, for election updates. Brandon Finnegan’s team is better and faster than any of the big media companies. I highly recommend their piece explaining what to look for as polls close in individual House races.
If you follow FiveThirtyEight and Nate Silver, you’ll know that his models are usually highly accurate. The probabilities he’s given The GOP for losing the House, and Democrats winning the Senate, are the same odds Trump had going into election day. So remember, as even he’d say, these are probabilities. Just because something has a 15% chance of happening doesn’t mean it never happens.
Oddities in the midterms
One odd thing in these midterms is that there hasn’t been a significant shift from one party to the other. Democrats held a pretty commanding lead in all vital polling data going into September, then the Kavanaugh hearings hit, and the polls narrowed. Most observers thought after Kavanaugh we’d see either a reversion to the mean or momentum towards one party or another.
In a typical wave election, what you’ll see happen late is all the undecided voters break for one party or another. If there’s enough of that group, it turns a good year for one party into a great year. That’s how waves get formed. Oddly enough, there’s been a sort of stasis effect on the polls since Kavanaugh’s impact waned and we had the news cycles of the immigrant caravan, the pipe bomb mailers, and synagogue shooting.
Instead, what we’ve seen is specific movement in races towards one person or another, but nothing uniform for one party. The move could happen this next week, but I doubt we’d see it. In the last week or so of campaigns, pollsters start fudging their numbers a tad and herd towards what everyone else is reporting (see this piece on poll herding in 2014). No pollster wants to be the odd man out, with a weird polling result that’s wildly out of line with the rest of the pack. So they tend to group, so failure seems systemic instead of an individual.
And when you stop and look one week away from the midterms, you see this odd stasis in races, with no momentum towards one party, and still room for undecideds or poll errors to swing the election. That uncertainty is what keeps the GOP with a puncher’s chance of maintaining power in the House.
The other odd piece of data is that if you poll battleground House districts, Democrat’s advantage evaporates. The Washington Post surveyed these areas and found that Democrats only had a 4 point edge on the generic ballot. They didn’t poll the President’s approval rating in this areas, but presumably, his approval rating should be higher too.
On the one hand, you can use the battleground poll as a way to show that things are tighter than national polls show. But you’re also missing out on the broader picture by only focusing on a handful of districts. But if Democrats hold a 4 point lead on the generic ballot, you could say right now Republicans are poised for a good night. Democrats tend to get a boost on the generic ballot due to their clustering in cities, which can boost an overall number, but not matter in suburban and rural districts.
Here’s why that generic ballot number is essential. Nate Cohn, of the New York Times, wrote a piece over the summer entitled: “A Democratic Blue Wave? Don’t Forget Republicans’ Big Hill.”
Today, there are just nine Republicans who represent districts that tilt toward the Democrats, based on the districts’ voting in the last two presidential elections compared with the country as a whole. There were 24 Republicans in such a precarious position in 2006, 67 such Democrats in 2010 and 90 Democrats in 1994.
In a wave election, Republicans representing Democratic-tilting districts like these would be projected to lose their re-election bids. They are the figures farther down the slope in the accompanying illustration. (The one all by itself at the bottom, representing Pennsylvania’s Fifth, was sunk by redistricting.)
Mostly, though, the districts are above sea level. Incumbents who represent even somewhat Republican-leaning districts are generally favored to win re-election, even in a wave election. Waves aren’t necessarily as deadly as you might think.
This makes it hard to imagine huge Democratic gains, like those the Republicans made in 2010. In past elections, incumbency has been worth around seven points, although that advantage has been declining.
Republicans have so many safe seats that the Democrats would be expected to gain only 22 seats if they flipped Republican-held districts at a rate equivalent to the waves of 2006 or 2010 (without factoring in the large number of open seats). It would leave them one seat short of a majority.
Why do Republicans have such a deep structural advantage in this era? For starters, a majority of districts lean Republican in presidential elections because of gerrymandering and the tendency for Democrats to waste votes in heavily urban areas. But there’s another factor: Most Democratic-leaning districts already have Democratic representatives, and most Republican districts already have Republican representatives.
In short, Democrats have all moved to fewer and fewer districts, leaving more area open to Republicans to poach. In my home state of Tennessee, this is especially true. The bulk of all Democratic votes comes from Nashville and Memphis, while the rest of the state is a deep red.
So the question is, are Democrats in big cities tilting things like the generic ballot and polling towards them? And if so, will that leave House seats without enough Democrats votes to move them?
It’s hard to tell because there’s been so little polling in important races. Most of the national media attention has focused on places like Ted Cruz in Texas, in his race against Beto O’Rourke (a race Cruz should win handily due to all the needless data we have on it).
Add to that Democrats relying almost exclusively on groups who typically don’t vote, like young voters and various racial groups.
So if turnout isn’t what Democrats need, and the polls are tilted ever so slightly because of Democratic herding into a few select districts, then you could get an instance of where the conventional wisdom is wrong.
In a nutshell, Democrats could be wasting votes in places that are already intensely blue. And because these areas are cities, it makes the marches and protests look impressive but also nullifies them electorally.
Conventional wisdom says GOP Senate and Democratic House. That’s the safe bet. It’s also likely wrong. Even though the Senate races are in mostly Trump country, I don’t see how one party wins one chamber while losing seats in the other. Polls and voters tend to move together. Republicans didn’t win the Senate in 2010, but they made significant inroads. If Democrats take the House, they’ll make inroads into the Senate. On the flipside, if the GOP has a good night in the Senate, it portends poorly in the House for Democrats.
My prediction: Republicans hold the House, and maintain a ten-seat edge, putting them at or near their current number in the House.
For the Senate, Republicans hold the Senate and have between 54-56 Senators expanding their current majority. Right now, neither McCaskill or Donnelly, both Democrats, are acting as if they’re leading their races. Arizona, Montana, and Florida are all harder to read. Florida, in particular, is hard to understand because Hurricane Michael could be throwing off the polls.
Tennessee and Texas will both swing easily towards Republicans. Phil Bredesen (D) has run a solid campaign, but I don’t see how he gets past the momentum of Bill Lee (R) on the Governor’s side of the race. And Cruz leads Beto in every conceivable way in the polls that I don’t see any path for Beto.
If I’m right, and Republicans have a better than expected night, one race to watch: Michigan’s Senate race. Senator Debbie Stabenow (D) vs. John James (R). Given Michigan’s blue history, and Stabenow running for her fourth term, she is strongly favored. The race didn’t receive much attention nationally (though it deserved more than Beto). But the few polls conducted did show an interesting trend. The Detroit Free Press polled the race showing Stabenow with a 23 point lead in September that had shrunk to 7 points in mid-October.
That dive caused some other pollsters to jump in late, showing Stabenow with a double-digit lead. But others are showing a single-digit lead as well. If true, it would be Stabenow’s stiffest test yet, against the far more charismatic and young John James (seriously, if you’re conservative, go check James out). It’s a pipe dream for me, but I’d love for James to win. And that odd poll shrinkage ended right when you’d expect poll herding to kick in gives me the slightest hope.
You have my predictions. Feel free to go over to RealClearPolitics and design your maps; it’s fun practice. If I’m wrong, we’ll find out in a week and go through the numbers and figure out what happened.
Links of the week
Satire piece of the week
U.S.—The nation’s liberals were “deeply grateful” that more and more strict constructionists have been appointed to both the Supreme Court and other federal court positions after President Trump suggested he would get rid of the concept of birthright citizenship as outlined in the 14th Amendment.
Although they fought the nomination of Kavanaugh and have otherwise criticized every move Trump has made in regards to the judiciary, they had to admit they were glad that justices who interpret the nation’s supreme document in a literal manner would be able to block any attack Trump made on the Constitution.
“It’s a good thing Trump has appointed judges who interpret the Constitution in such a narrow way now,” said one woman in Oregon. “If we’d gotten judges up there who didn’t care about the Constitution one way or another, who knows what they would have ruled.”
Bonus – The Onion brings back Joe Biden: Biden Hands Out Loose GT Cola Can To Unexpected Trick-Or-Treater
Thanks for reading!