2018 Congressional Generic Ballot and House Polls Megathread - the original

[Devout Centrist]

YouGov, May 6-8, 1228 RV

D 44 (+2)
R 35 (-4)

Crosstabs indicate that Democrats are winning in the Northeast (Census region; really New England +Mid Atlantic) 48-31(!), in the Midweast 43-34, and in the West 48-33(!).

IF this is close to what the final margins look like in November, it's very hard to see Republicans holding the House.

[Kodak]

I don't think what Trende is saying is necessarily bad or wrong or anything. He's just going off the data he has, and without a doubt the data he is using doesn't portray a Democratic wipeout (although like he implied, still a sizable win that could flip the House).

I think the difference between him and some of us is that many of us are looking at where we think the race will be vs where it is right now. Also if you weigh the special election results more heavily, it does give you a reasonable basis for believing that the gcb polls aren't necessarily reflecting what the environment is and/or will be in November. I think there is a pretty reasonable chance that the floor falls out from under Republicans when September/October comes around. I think that would be somewhat in sync with past wave elections, where some of them broke late in the cycle.

There is also the PPP argument (which isn't just theirs, really) that the environment is actually stable and more in Democrats favor and the polls are just bouncing around right now as people play footsie with the idea of supporting someone they won't actually support come election day. I don't know how to prove this though.

Obama had 46% approval in 2010, and we know what happened.

To be fair, Democrats were severely overextended in 2010. Republicans aren't in a similar position, which is why a ~6.8% win for Democrats wouldn't flip over 60 seats.

It does't necessarily require the incumbent president to be super unpopular to generate a backlash. Generally the cutoff I've seen is 50%. If the approvals are under 50%, wild things can happen, and there might not be a significant difference between 39% and 44%.

The Republicans aren't overextended at the federal level, but they certainly are at the state level, even more than the Democrats were in 2010, and that's bad. I suspect that having a lot of incumbent senators from the opposition party and a lot of incumbent governors from the party in control on the same ballot is the worst possible combination for the party in control.

[Virginiá]

The Republicans aren't overextended at the federal level, but they certainly are at the state level, even more than the Democrats were in 2010, and that's bad. I suspect that having a lot of incumbent senators from the opposition party and a lot of incumbent governors from the party in control on the same ballot is the worst possible combination for the party in control.

Yes, that's true. The electoral landscape at the state level has shifted a pretty good deal under Obama, except Republicans have been able to hold on to seats they would otherwise lose later on because practically every election under Obama was either a Republican wave or only a lightly favorable year for Democrats (2012). This allowed Republicans to snatch up a bunch of offices that they wouldn't otherwise hold, at least long-term. Very curious to see what the situation in the states is come Jan 2019.

[Dr Oz Lost Party!]

As always, I got the cross tabs

The GOP's Midwestern collapse is REAL. The Dems are winning there 43%-34% Rod Blum and Erik Paulsen are finished

18-29 year-olds going for Dems 58-21%

Blacks going 73-8% in favor of Dems

Hispanics only give the GOP 20% in support, 53% for Dems

Dems winning moderates 49-21%

[ON Progressive]

18-29 year-olds going for Dems 58-21%

But I was told millennials are leaving the Democratic Party!

[Dr Oz Lost Party!]

Lol The Federalist is celebrating the fact the GOP's black support is 8% in the YouGov poll

https://twitter.com/wfbtfm/status/996145645747671040

[Dr Oz Lost Party!]

Interesting, RCP has the YouGov poll D+9 (5/6-5/8) entered in as if it was taken BEFORE the CNN poll D+3 (5/2-5/5)

Is this a mistake or is "Sean T at RCP" trying to pull one over us to try and keep his "DEMS ARE DOOMED" theory alive?

[Darthpi – Anti-Florida Activist]

As always, I got the cross tabs

The GOP's Midwestern collapse is REAL. The Dems are winning there 43%-34% Rod Blum and Erik Paulsen are finished

18-29 year-olds going for Dems 58-21%

Blacks going 73-8% in favor of Dems

Hispanics only give the GOP 20% in support, 53% for Dems

Dems winning moderates 49-21%

I think my favorite statistic in there is Democrats winning among people who are "Not sure" what their ideology is 31%-2%, with the rest either undecided or not voting.

[KingSweden]

I’m not going to dig too deeply into the cfosstabs of a poll as jumpy as YouGov. +6 in a week? Who knows what it’ll be next week.

[IceSpear]

I’m not going to dig too deeply into the cfosstabs of a poll as jumpy as YouGov. +6 in a week? Who knows what it’ll be next week.

Or in 26 weeks.

[KingSweden]

I’m not going to dig too deeply into the cfosstabs of a poll as jumpy as YouGov. +6 in a week? Who knows what it’ll be next week.

Or in 26 weeks.

Yes, indeed.

[LimoLiberal]

Cook Political Reporter implies that the political environment is shifting fast.

https://twitter.com/amyewalter/status/996355701474840576

[Hindsight was 2020]

F**k off Limo seriously this bit is getting really obnoxious

[LimoLiberal]

F**k off Limo seriously this bit is getting really obnoxious

If Amy E Walter had tweeted "Democrats set for massive wave. Trump is worst president of all time." nobody would care if I posted that. Walter isn't some Republican hack. She's experienced, and I trust her insight especially because she has been relatively bullish on prospects of a Democratic wave recently.

[Wisconsin SC Race 2019]

F**k off Limo seriously this bit is getting really obnoxious

Dispute the fact and argue on that, instead of cussing at him.

[Hindsight was 2020]

Oh spare me Limo the biggest political needs like G. Elliot Morris have said that this is suppose to be the best time polling for the party in the WH during the midterms. The fact that dems are leading by around the 6-7 points they have since forever is better than the reps were in 2010 and 2014.

[Gass3268]

At the same time that Trumps approval has improved by about 4 points, the GCB has stayed relatively flat. 538 has seen the GOP gain about a point while HuffPost Pollster has actually seen the Dems gain about a point. It's hard to compare to RCP as they only use the 5 most recent polls and not the totality of all recent polls like 538 and Pollster does.

[KingSweden]

At the same time that Trumps approval has improved by about 4 points, the GCB has stayed relatively flat. 538 has seen the GOP gain about a point while HuffPost Pollster has actually seen the Dems gain about a point. It's hard to compare to RCP as they only use the 5 most recent polls and not the totality of all recent polls like 538 and Pollster does.

I’d be curious how much of this is GOPers who have come home to the President, but who detest Congress (always remember how low Mitch and Paul’s ratings are).

[Hindsight was 2020]

Benchmark Model released their first House Model which has the dems winning 26 seats https://shareblue.com/benchmark-model-gop-lose-house-midterms/amp/

[Holy Unifying Centrist]

Benchmark Model released their first House Model which has the dems winning 26 seats https://shareblue.com/benchmark-model-gop-lose-house-midterms/amp/

These models are still pretty meh. PA-17 being Lean R? I'm pretty sure if dems don't win PA-17, they aren't winning 15 house seats. Conor Lamb should be able to win this race easily.

[Gass3268]

Benchmark Model released their first House Model which has the dems winning 26 seats https://shareblue.com/benchmark-model-gop-lose-house-midterms/amp/

These models are still pretty meh. PA-17 being Lean R? I'm pretty sure if dems don't win PA-17, they aren't winning 15 house seats. Conor Lamb should be able to win this race easily.

They flat out say in their analysis that they are probably underestimating Lamb.

It's good to see more models! I'll have to add it to my aggregate house model when I have time.

[Hindsight was 2020]

Benchmark Model released their first House Model which has the dems winning 26 seats https://shareblue.com/benchmark-model-gop-lose-house-midterms/amp/

These models are still pretty meh. PA-17 being Lean R? I'm pretty sure if dems don't win PA-17, they aren't winning 15 house seats. Conor Lamb should be able to win this race easily.

Agreed so it's quite telling the dems still win the house with a bit of an R friendly model

[Hindsight was 2020]

The political environment is changing, part 2.

https://twitter.com/JakeSherman/status/996437200014278662

CLF polling in the 48 most competitive congressional districts has D+3. It was D+11 in December.

For God sakes Limo you are so desperate to annoy you are citing a GOP Superpac Internal?

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Apparently the political environment changes rapidly.



Next up, in the evening we over-react to the primary results.

[Blackacre]

That was 10 points a little while ago. Something's happening...