Current House Rating?

[UncleSam]

Tilt D

D+25

I still think Dems will have a lot more trouble taking Trump districts than people here think. Dems hold one Trump-won district in Virginia right now, but he won 235 districts nationwide. I just doubt that when it comes down to marginal Trump districts Dems will totally sweep

That being said, Dems have a huge generic ballot advantage and I expect them to sweep the Clinton-Republican seats, of which there are 24. I expect they'll more or less win 20 of those, 5-10 Trump districts, and Rs win a few districts Trump won big like MN-1.

Ds in popular vote by 7-8.

[smoltchanov]

Tilt R. D+20

[Oryxslayer]

Tossup. I need to see more retirements from seats that areor under R+4. Currenty there are 5: FL-27, NJ-02, WA-08, MI-11, PA-15, and a sixth in AZ-02 is probably on the way. Dems probably need a good 5-6 more. The magic number for me is 1/3 of seats needed to flip the chamber + half the number of similar PVI open Dem seats - currently that magic number is 10.

[Pessimistic Antineutrino]

Tilt R. D+20

Pretty much this, but I only envision it getting worse for Rs.

[Brittain33]

For people who do not see Dems taking over the House, I am curious what circumstances you would consider required to believe Republicans would lose the House. It is hard for me to envision the situation getting much worse than this for Republicans, but it's also hard for me to imagine Republicans being the majority party in the House of Representatives for eternity.

[Holmes]

For people who do not see Dems taking over the House, I am curious what circumstances you would consider required to believe Republicans would lose the House. It is hard for me to envision the situation getting much worse than this for Republicans, but it's also hard for me to imagine Republicans being the majority party in the House of Representatives for eternity.

I think with almost 8 years in power in the House, Republicans maintaining control has just become the default for so many people. Kind of like how 1994 was so surprising for the opposite reason, even though the signs were there.

[Pessimistic Antineutrino]

For people who do not see Dems taking over the House, I am curious what circumstances you would consider required to believe Republicans would lose the House. It is hard for me to envision the situation getting much worse than this for Republicans, but it's also hard for me to imagine Republicans being the majority party in the House of Representatives for eternity.

It really comes down to the seat math. The circumstances for Republicans to lose the House are there and are probably about as bad as it gets, but 24 pickups is still a tall order. I am definitely trying to be cautious here rather than forecasting D+30 without knowing numerous paths to that number. No one party ever wins every single tossup race, even in 2006 or 2010, and it won't be enough to have 30 tossup races because Democrats most likely won't win all of those. By "getting worse for Rs" I mean in terms of more and more retirements like Reichert and LoBiondo, which I expect in the coming months. Once it becomes clear that those types of seats are Safe D and people like Denham, Knight and Issa are not only in trouble but DOA, then I will forecast Dems to win the House. I keep it Tilt R out of caution more than anything.

[Brittain33]

For people who do not see Dems taking over the House, I am curious what circumstances you would consider required to believe Republicans would lose the House. It is hard for me to envision the situation getting much worse than this for Republicans, but it's also hard for me to imagine Republicans being the majority party in the House of Representatives for eternity.

It really comes down to the seat math. The circumstances for Republicans to lose the House are there and are probably about as bad as it gets, but 24 pickups is still a tall order. I am definitely trying to be cautious here rather than forecasting D+30 without knowing numerous paths to that number. No one party ever wins every single tossup race, even in 2006 or 2010, and it won't be enough to have 30 tossup races because Democrats most likely won't win all of those. By "getting worse for Rs" I mean in terms of more and more retirements like Reichert and LoBiondo, which I expect in the coming months. Once it becomes clear that those types of seats are Safe D and people like Denham, Knight and Issa are not only in trouble but DOA, then I will forecast Dems to win the House. I keep it Tilt R out of caution more than anything.

But what about the generic ballot polls ranging from +10 (Morning Consult, most pro-R poll) to an exuberant +18 on CNN?

[smoltchanov]

For people who do not see Dems taking over the House, I am curious what circumstances you would consider required to believe Republicans would lose the House. It is hard for me to envision the situation getting much worse than this for Republicans, but it's also hard for me to imagine Republicans being the majority party in the House of Representatives for eternity.

It really comes down to the seat math. The circumstances for Republicans to lose the House are there and are probably about as bad as it gets, but 24 pickups is still a tall order. I am definitely trying to be cautious here rather than forecasting D+30 without knowing numerous paths to that number. No one party ever wins every single tossup race, even in 2006 or 2010, and it won't be enough to have 30 tossup races because Democrats most likely won't win all of those. By "getting worse for Rs" I mean in terms of more and more retirements like Reichert and LoBiondo, which I expect in the coming months. Once it becomes clear that those types of seats are Safe D and people like Denham, Knight and Issa are not only in trouble but DOA, then I will forecast Dems to win the House. I keep it Tilt R out of caution more than anything.

+100. Exactly my thoughts too. I would only add, that Democrats effectively waste an enormous number of Democratic votes in 80%+ urban districts, so, even with clear general preference of public for Democratic candidates now, the preference in swingy and slightly Republican districts (where most turnover must happen, i think no one thinks that Democrats will win TX-13 or AL-06) is, most likely, substantially smaller. Republicans have less "overwhelmingly Republican" districts then Democrats - "overwhelmingly Democratic", but - substantially more districts, "leaning" in their direction (usually) then Democrats. And second, preference for "generic Democrat" doesn't always correlate well with preference for specific candidate. People can generally prefer Democratic candidate for their district, but "this one" specific candidate can rub them the wrong way. So, polling for "generic" party candidate is, frequently, better then for real candidate.

[Brittain33]

True, but if Democrats waste those votes in districts that already vote 90% Democratic... is it possible that the big swings in the generic ballot we are seeing are necessarily coming from less "packed" districts? Dems may not be maxed out in the urban districts because turnout was low in 2016, but it may well be that we will see larger swings in the moderate and R-leaning districts precisely because Dem districts were maxed out before the anti-Trump bump.

[smoltchanov]

True, but if Democrats waste those votes in districts that already vote 90% Democratic... is it possible that the big swings in the generic ballot we are seeing are necessarily coming from less "packed" districts? Dems may not be maxed out in the urban districts because turnout was low in 2016, but it may well be that we will see larger swings in the moderate and R-leaning districts precisely because Dem districts were maxed out before the anti-Trump bump.

May be. And still - if, generally, Democrats have 10-18% advantage according to most polls then, naturally, in these very urban districts they may have 50-60% (and more) advantage. And because there are more such "uber-Democratic" districts then "uber-Republican" - these percentages are not completly cancelled by these "uber-Republican" districts. So, in other districts (considered separately), percentage will generally be lower.

An example of the same phenomenon: IIRC Hillary won over Trump by about 3%. But Trump won 230 districts, Hillary - 205. Why? Because of greater number of "super-Democratic" disitricts with astronomical percentages for Hillary and, generally, Democratic candidates. These votes go into general calculation, but district distribution still favors Republicans.

[smoltchanov]

True, but if Democrats waste those votes in districts that already vote 90% Democratic... is it possible that the big swings in the generic ballot we are seeing are necessarily coming from less "packed" districts? Dems may not be maxed out in the urban districts because turnout was low in 2016, but it may well be that we will see larger swings in the moderate and R-leaning districts precisely because Dem districts were maxed out before the anti-Trump bump.

May be. And still - if, generally, Democrats have 10-18% advantage according to most polls then, naturally, in these very urban districts they may have 50-60% (and more) advantage. And because there are more such "uber-Democratic" districts then "uber-Republican" - these percentages are not completly cancelled by these "uber-Republican" districts. So, in other districts (considered separately), percentage will generally be lower.

An example of the same phenomenon: IIRC Hillary won over Trump by about 3%. But Trump won 230 districts, Hillary - 205. Why? Because of greater number of "super-Democratic" disitricts with astronomical percentages for Hillary and, generally, Democratic candidates. These votes go into general calculation, but district distribution still favors Republicans.

Wtf that kinda proves that a 12% win on the GCB ballot would flip the House. Hillary won the PV by 2 points and would only need to win 13 more districts to have gotten a majority. 12% PV win would definitely finish the job and then some.

It proves nothing. A 12% PV win can consist of 50% PV win in already Democratic districts (zero flips), 20% loss in solid Republican districts (again - zero flips), and, say, 2% win in swingy districts (with modest number of flips), which are THE most important in seats distribution.

[Virginiá]

I just don't see a reason to doubt the generic polls. There have already been a number of convincing analyses by 538, Sabato and others regarding this. They found that the average of the generic polls this far out is actually a lot more predictive than people think, and that things generally get worse for the president's party. Further, a recent Sabato article even argued that the magic number for the House PV may actually be lower - around 4 or 5 points I believe, although I'm not sure if that's right, but we'll see.

My thought is that if Democrats win the House PV by 9+ points, it doesn't matter what each race's state (tossup/etc) is. The House is likely gone if that is the House PV. That is simply too much for Republicans to absorb imo. They have favorable House maps but it's not that favorable. If it's <9 points then I'm a little more hesitant to just call it.


Leans D for now, but if things are still just as bad (give or take) by March or so, Likely D.

[Virginiá]

It proves nothing. A 12% PV win can consist of 50% PV win in already Democratic districts (zero flips), 20% loss in solid Republican districts (again - zero flips), and, say, 2% win in swingy districts (with modest number of flips), which are THE most important in seats distribution.

smoltchanov You're talking theory here but you know that is not going to happen. Name one election where a party won a huge popular vote advantage but saw all of it get sunk into unwinnable districts, and thus didn't win nearly as many seats as it would suggest. It doesn't happen. The vote distribution is a lot more even in that.

[smoltchanov]

It proves nothing. A 12% PV win can consist of 50% PV win in already Democratic districts (zero flips), 20% loss in solid Republican districts (again - zero flips), and, say, 2% win in swingy districts (with modest number of flips), which are THE most important in seats distribution.

smoltchanov You're talking theory here but you know that is not going to happen. Name one election where a party won a huge popular vote advantage but saw all of it get sunk into unwinnable districts, and thus didn't win nearly as many seats as it would suggest. It doesn't happen. The vote distribution is a lot more even in that.

Of course i present a morst extreme case. Just as it's possible theoretically to win 49 states by 1 vote each, lose 50th by million, and have overwhelming absolute majority in electoral college as a result. Real distribution is somewhat more uniform, but - still far from being uniform. And, as i said - there are considerably more uber-Democratic districts, then uber-Republican. So, Democratic advantage in swingy districts alone is, probably, about, say, 6%, not 12% (overall for country), the other 6% - coming from that big majority in safe districts and their relatively big number. This may be enough too, but still - there are variants here.

[smoltchanov]

It proves nothing. A 12% PV win can consist of 50% PV win in already Democratic districts (zero flips), 20% loss in solid Republican districts (again - zero flips), and, say, 2% win in swingy districts (with modest number of flips), which are THE most important in seats distribution.

smoltchanov You're talking theory here but you know that is not going to happen. Name one election where a party won a huge popular vote advantage but saw all of it get sunk into unwinnable districts, and thus didn't win nearly as many seats as it would suggest. It doesn't happen. The vote distribution is a lot more even in that.

And frankly there's not too much more for Dems to juice out of many of these Dem vote sinks. Not to mention the largest swings almost always happen in competitive seats, as several studies have found

Not only swing size, but absolute number and majority of advantage in non-swing districts is also important.... I agree that situation is very favorable for Democrats NOW, and if it remains this way by October - i will change my prediction to "Lean D" (at least). But last midterms of 2010 and 2014 convinced me, that Democrats usually underperfom their initial expectations (sometimes - severely) - hence my cautious forecast for now.

P.S. Brittain33 asked for a reasons i (and similar people) are so cautious in our predictions. I presented them. One can agree or disagree, but explanation is given at least..

[IceSpear]

Tilt D

D+25

I still think Dems will have a lot more trouble taking Trump districts than people here think. Dems hold one Trump-won district in Virginia right now, but he won 235 districts nationwide. I just doubt that when it comes down to marginal Trump districts Dems will totally sweep

That being said, Dems have a huge generic ballot advantage and I expect them to sweep the Clinton-Republican seats, of which there are 24. I expect they'll more or less win 20 of those, 5-10 Trump districts, and Rs win a few districts Trump won big like MN-1.

Ds in popular vote by 7-8.

This, more or less.

[King Lear]

Lean R
D+10
I still believe republicans are favored to hold the house due to the simple fact there are 230 Trump districts
to 205 Clinton districts, as we saw in the Virginia house of delegates races were Democrats gained at least 15 seats, they struggled to win in trump districts and had to sweep as many Clinton districts as possible, However I cannot rule out a scenario where democrats manage to flip the house though I do not believe it is likely at the moment, hence my Lean R rating.

[Chief Justice Keef]

Lean/Likely D, Dems gain 50 seats.

[Pericles]

The biggest swings to Democrats in the special elections were in solidly Trump districts like MT-AL, KS-04 and SC-05, and of course Alabama. It's entirely possible and indeed probable that Democrats will also make improvements with the WWC in 2018.

[Gustaf]

I'd say Lean D. It is clear they would win the House with the current GCB lead. There isn't much reason to assume it would improve for Republicans either. We're still sufficiently far away from election day that I don't think it makes sense to go more than Lean D but I really don't understand why someone would think it is favoured to remain Republican.

In particular I don't think it is at all clear that the SWING will happen more in solid D districts. Like, maybe but I haven't really seen any evidence suggesting that. What little I've seen seems to suggest the opposite as more likely.

[Amenhotep Bakari-Sellers]

Dems will win between 15-25 seats as of today.  But, that will change, I say tilt D pickups in CA, TX, FL, WI, MI and PA

[MAINEiac4434]

Likely D, closer to safe than lean.

[King Lear]

Likely D, closer to safe than lean.

The overconfidence is mind blowing, you’d think people would learn from last year.

[Nyvin]

Likely D, closer to safe than lean.

The overconfidence is mind blowing, you’d think people would learn from last year.

Learn from 2009 would be more like it.