Mid-term prediction

Politics & Elections
1

It’s 10 months until the mid-terms. Time for some predictions.

Currently the House of Representatives has 239 Republicans, 193 Democrats, and 3 vacancies. The vacancies in Pennsylvania and Arizona will be filled through by-elections in the next few months; the Michigan vacancy will stay open until the November election.

So lets predict how many seats will change hands in the mid-term election. All you need to do here is predict how many seats will change between now and after the election. I’m not asking about specific districts, just the overall number. If you think the Dems will pick up 25 seats, just say D+25; for the same prediction for the GOP: R+25. If you want to predict changes in the Senate, Governorships, or state legislative houses, feel free to make those too. You can explain your prediction(s) if you want, but don’t have to.

OK, so my prediction: D+100.

I’m going to spoiler my explanation in case anyone doesn’t want to be influenced.

[spoiler]
That’s rather extreme (although I consider it to be conservative; a D+125 would not surprise me). I don’t think such a large shift has ever happened before, but these are not ordinary times. I base my prediction on two factors: Democrat base being activated by the current administration and gerrymandering.

The first factor is going to bring out Democrat voters who rarely vote in the mid-terms. They’d like to vote against the president but are frustrated by the 4-year term, so they’ll come out to send a message. A rather large percentage of swing voters (probably a majority) will want to send the same message. The Republican base is not going to be energized like this at all. In fact, some may be energized to send the same message as the Dems.

The second factor may seem paradoxical, but it comes from a potential weakness in gerrymandering. Generally, gerrymandering favors the party doing it by giving it a surplus of voters in the majority of districts. But that majority is generally rather slim in most districts, since the limited number of party voters have to be spread among many districts. Normally, this doesn’t make any difference. About the same percentage of voters from each party will vote in a given election, so the advantage holds in most elections.

But this isn’t an ordinary election. Because of factor #1, the Dems should have a much higher percentage of their voters turn out, which will overcome the normal disadvantage they have through gerrymandering. This should be decisive and vote out many Republican incumbents.

Assuming that the Democrats get some quality candidates, that is. But I think factor #1 should also produce this. People who wouldn’t normally consider running are going to file. There should be good candidates among them.[/spoiler]

So what’s your prediction for the House next session?

2

I’ve read your explanation, and I think you’re being far too optimistic. Without doing a district-by-district study, I know that there are too many solidly-red seats for your model to work out and too many people who will absolutely never vote for anyone with a (D) after their name.

Forget 100 seats, which has happened once (during Grover Cleveland’s term in 1894, so not exactly what would be called ‘modern era,’ when Republicans won). I don’t have any confidence in the electorate taking back even the 30 seats they’d need for control of the House. Too many of P.T. Barnum’s suckers around.

3

Part of my point was that a lot of those people won’t vote in the mid-term. They aren’t energized like D’s are. So what’s your prediction?

4

Yet R’s more regularly vote in midterms.

As for my prediction, I’'d go with D+20 or so. But that’s a complete guess.

5

Consider this scenario: R’s vote in their usual mid-terms numbers, D’s vote in Presidential-election numbers, and Unaffiliateds break 2-1 or better for D. How many districts are still solidly red then?

We saw something like this in the Alabama senate election. That was unusual for another reason, so maybe we can’t count on that as a good bellwether. The two upcoming by-elections will give a better idea, especially since, AIUI, the Arizona one is a very red district.

6

dtilque, I agree with you. I’ll go with D+25 and maybe as much as 80 – though I wouldn’t be surprised if it’s more.

My reasoning spoilered as yours was:

[spoiler]1. Dems are fired up. And not just Dems. Moderate Republicans and independent voters, too. They are going to vote in the mid-terms in presidential election year numbers in 2018. Minorities and women will turn out in record numbers as well.

  1. There are going to be a lot of new voters participating, some for the first time in their lives. They’ve gotten a wake-up call that democracy is not a spectator sport, and they’ll come out to vote in this election.

  2. The number of people – women especially – who are running as Democratic candidates for local, state and national congressional positions is jaw-dropping. The number is off the charts from any election in recent history. Dems mean to run someone in every single district in the country and there is no shortage of recruits. At the same time, Republicans are suffering both a recruitment problem and an attrition problem. Lots of House Republicans are leaving office after their terms are up in 2018. All this means that Dems will run unopposed in a lot of districts. Dems will win by default and the party won’t even have to spend a dime for it to happen.

  3. Serious enthusiasm gap between Dems and Republicans.

  4. It’s not going to be a money election. It’s going to be an ethics election.

  5. The customary advantage given to the out-party in mid-terms will be significantly amplified in 2018. To this add a generic ballot that favors Dems right now by 13 points – and likely by more nearer November, owing to continuing information about corruption revealed by Mueller’s investigation and the continuing nutty antics of the Trump administration.

  6. More and more, people are becoming aware of the Russian interference in 2016 and how it was implemented. Forewarned is forearmed. Fewer people will be vulnerable to such memes and fake news they read on social media in 2018.

  7. Numbers coming out in special elections across the country are consistently favoring Dems by roughly 20 points. It’s not just an aberration from only one or two elections.

  8. Just because Bannon has been sidelined doesn’t mean the wing nut faction of the Republican party won’t offer primary opponents in vulnerable districts where they think they can win. That means Trumpite candidates like felon Michael Grimm will run in some House districts, taking resources away from the McConnell Republican candidates just to get to the general election. Trumpites appear to have far more trouble with this division within their party in the Senate races, however. (Yayy!!)

  9. Dems and progressives have figured out how to run their races, as demonstrated by Doug Jones in Alabama. Focus on kitchen table issues, run your local race and don’t worry so much about the purity test crap. No progressive’s vote in such a race is going to be stolen over to the Trump camp, and neither will any Dem’s.

  10. People are shit scared of this “president.” They want to restrain him through every avenue available. Voting is an easy one.[/spoiler]

7

I deleted your spoilered stuff, but I’m going to comment on some of it anyway.

I think you’re way too optimistic in your reasoning. Remember, we’re talking about roughly 35% approval ratings, which is pretty much Trump’s floor, no matter what stupid shit he pulls. And that’s almost completely made up of Republicans and people who are either blind or completely un- or mis-informed.

Now that 35% aren’t evenly distributed, and they’re generally among the most active voters. They obviously tend to live in the red districts, which means their election results are unlikely to be any different than the last ones, regardless of who the Dems find to run.

Jones is a special case; Moore is such a sleazeball that Alabamans had little choice but to vote against him or look like complete idiots. As it turned out, it was still close enough that Moore could fight it. How many other congressional races are going to attract that kind of national attention, let alone have such an obvious villain?

8

With all those reasons, I’m surprised you only went with D+25. That number is roughly the average midterm swing. Forty-plus seat shits are not that uncommon. This is not a typical election, so I expect an atypical result. Hence my D+100 prediction.

9

History alone suggests that mid-terms usually favor the party in opposition to the president, so it seems likely that the Democrats will make some gains.

In the case of this particular president, however, he is energizing an entirely new kind of political energy. There are not only more motivated voters coming out to the polls, but he’s encouraging people to become politically active, particularly in local elections. The question in my mind is when does this tidal wave of opposition crest? It’s clearly building, and it’s clearly having an impact on local races, but does it build in time to change the congressional landscape for November of 2018?

The next several months are going to be pivotal for both parties. The Democrats have the advantage but they could still squander it. Tax reform, jobs growth, and even wage growth may also end up favoring Republicans in the short term. I would caution the Democrats not to force a shut down over DACA - do it over CHIP and healthcare reform but not DACA. That’s a loser.

As hyperbolic and Chicken Little-ish as it sounds, my concern is just having elections that are viewed as legitimate. The investigation of collusion between the Republicans and Russia is going to continue to have real-world consequences and threaten their agenda. And as it continues, it threatens their grip on power. These people don’t play by rules. Vladimir Putin has a very clear motive to continue disrupting America’s elections, and the Republican party - the odds-on favorite to lose the elections - have a very clear motive in receiving a “providential” hand in helping them remain in power. I have a bad feeling that we’re going to see something we’ve not seen before.

10

Still a ways away and things could happen to upset the apple cart here. Things in the Democrats’ favor: People’s impressions of Donald are pretty baked in. You’ve got the 25% or so who would happily lap up the sweat dripping from his balls, they aren’t going to change. Neither are the 25% who wouldn’t piss on him if he was on fire. Those in the middle are seeing for themselves that he is incapable if not insane. A lot of these folks are going to vote against him even if he isn’t on the ballot.

Something I think is often overlooked: Gerrymandering works out great for the party doing it in normal times, but they set themselves up for disaster if there is a wave election. Yes, you give up a few 90-10 seats in order to get a shitload of 55-45 seats for yourself. But if your party shits the bed big time, those 55-45 seats can flip against you. And I think that’s what’s going to happen.

D+80.

11

A lot will depend on what happens in those ten months, and in this chaotic environment it’s difficult to predict anything.

One factor to consider is that there are 14 Republican Congressmen from California, a state that just got royally screwed by the recent tax bill. Issa, Nunes, Rohrabacher - all their seats could be knocked further into play by disgruntled constituents (and frankly the three of them can fuck right off IMO). The Democrats only need, what, 24 seats to flip to take the House? Those 14 seats could make that a much more easily achieved goal.

I think D+100 is ludicrously optimistic but if (and this is a big if) the Democrats can get out the vote properly D+25 is certainly doable and D+50 potentially so. And the recent Alabama vote already provides evidence to Dems that they can win against long odds if they turn out.

12

I think any predictions this early are essentially guesses. For an actual prediction, I’ll predict that Nate Silver’s day-before prediction will be very close to the actual result.

For my guess, I’ll say Dems +40.

13

If I were in charge of the DNC, the best strategy I could think of would be to use the Republicans’ strategy and ideology against them. You want states’ rights? You think power should be stripped away at the federal levels and that states should be more powerful? Okay, then let’s play that game and let’s see Democrats start investing more of their resources into state and local elections, which is where most of the power that impacts people on a daily basis rests anyway and will for the foreseeable future. One of the biggest problems that Democrats have had is that they keep trying to hit home runs in federal elections. Get guys on base first (i.e. meaning start racking up wins in smaller races that may not be so important nationally but become important if they win enough of them).

14

I’m also going to go with “hopefully” D+100, but for different reasons.

The current generic ballot on 538 (and RealClearPolitics) has been showing +10 to +15 for Democrats for a few months now.

If we leave some room for the “incumbent advantage” (say 1/3, down to +10), the Cook Partisan Voter Index has 114 seats that are currently held by Republicans that have PVIs of R+10 or less (meaning they could be a Democratic district through a Republican district by up to +10).

The generic ballot historically hasn’t changed significantly between Jan 1 of an election year and the election, combined with the other reasons noted above around Democratic voter enthusiasm, Trump’s historically low approval rating, etc., make me hope this wave will be historically enormous.

There’s no reason why it shouldn’t - there are so many other abnormal things happening in Washington, a triple-digit swing seems very possible to me right now.
Some other quick numbers of Republican held seats based on PVI:

R+9 – 104
R+8 – 84
R+7 – 72
R+6 – 62
R+5 – 52
R+4 – 43
R+3 – 32
R+2 – 25
R+1 – 18

We’d need to get to only R+1 for the Dems not to retake the House this year, which seems … unlikely.

As much as I hope we’re in the R+10 range and we’re talking 100+ seat swing, if I had to bet, I’d also wager in the 45-55 range to flip.

15

Five of those 14 districts favored the Republican by more than 25% in the last election, including the CA 22nd for Nunes, who was at +36.4%. I’ll go out on a limb here and say that none of them is going to swing blue this year.

Another half-dozen districts went red by between 15% and 20%, including Rohrabacher’s. Again, not much chance of flipping any of them.

Issa is by far the most vulnerable; he won by a mere 2%, the smallest margin for any California House seat. And two other seats were won by Pubbies with margins of 4.8% and 8.4%.

Unless that blue wave is of epic proportions, I don’t see Dems gaining any more than three CA House seats this year…which also assumes that they hold on to all of the other seats. (There were three Dem seats where the margin was under 10%.)

16

I’m going to guess D+32. No, +33. Mark it down!

17

I actually made that point in my spoilered reasons, but it doesn’t hurt to repeat it. And I think you phrased it better.

By all rights, this election should be similar to Canada 1993. You may remember that one where the Progressive Conservatives went from a majority (156) in Parliament to just 2 seats. It won’t be that extreme, but only because there’s no significant third party around to take away the R’s core voters.

18

You’re looking at it from a macro level when it really isn’t determined that way. Sure, you can look back with 20/20 hindsight once the election is done and try to figure out why things went the way that they did. But as I demonstrated in CA, a sizable swing to the Dems would likely only result in a gain of 3 seats. Granted, it’s an already-blue state, but also has the largest number of seats.

We can throw around numbers all day long, but they don’t mean much. So I ask where, exactly, are those 100 seats supposed to come from?

19

Looking at the races on 270towin.com: If Dems hold all 174 “Safe Dem” seats, grab all 17 “likely Dem” and “lean Dem” seats, plus win all 20 toss-up races, that puts Dems at 211 House seats-- 7 shy of a majority. (And that’s a big, big “if.”)

So if we give Dems 25% of the “lean Repub” and 5% of the “likely Repub” (hey, perhaps we’ll see a few more upsets), that gives the Dems the 7 they need to take the gavel.

That means a D+27. That’s my prediction.

20

You’re right, I am looking at it from a macro level. I’m not going to get to do a micro level analysis. That would be a waste of time, since it would be based on ordinary election results. The point is, this isn’t going to be an ordinary election.

So they’re going to come from districts that are normally safe red districts, and that will mostly go back to being safe red districts. But this one time, or in 2020 too if Trump runs for reelection, people who don’t normally vote are going to come out to vote D. Or more specifically, to vote against Trump, but the best they can do for this election is vote against a Republican congressman and Senator. And not just them. Unaffiliated voters are going to be doing the same. While they may usually vote about the same way as the affiliated voters in their district, this time they’ll break 2-1 or better for the Dem candidate. Even in safe red districts.