2018 Congressional Generic Ballot and House Polls Megathread - the original

[GeorgiaModerate]

https://twitter.com/Nate_Cohn/status/994313554504404998

[GeorgiaModerate]

http://observer.com/2018/05/jeff-van-drews-internal-poll-has-him-leading-hirsh-singh/

Van Drew up in NJ-02 by 18 points, 52-34.

Knock off 5 for an internal, but that's still pretty impressive.

[GeorgiaModerate]

https://twitter.com/gelliottmorris/status/994975779774783488

Any indication of which polling firm they used, for database purposes? I wasn’t aware Kaiser did this kind of thing in-house, but who knows

They used SSRS.  I think this is the poll in question: https://www.kff.org/health-costs/poll-finding/kaiser-health-tracking-poll-preview-role-of-health-care-2018-midterm-campaigns/.  From the Methodology link:

You must be logged in to read this quote.

[GeorgiaModerate]

The only information I can find in the poll release about which races are battlegrounds is this:

You must be logged in to read this quote.

[GeorgiaModerate]

Ipsos/Reuters, May 6-10, 1317 RV

D 39 (nc)
R 37 (-1)

[GeorgiaModerate]

People’s Pundit Daily is restarting Trump approval polls/generic ballot polling next week.

Anyone who posts this crap gets an auto ignore from me.

What is people pundit daily?

A right-wing news site.

[GeorgiaModerate]

Just popping in to say that Democrats lead in the generic ballot in May 2010. In fact, they lead as late as July 2010 (and as late as September 2014.)

Didn't the Democrats actually win the popular vote anyway, despite not winning the House? The 2010 "election" was based on the 2000 census, which had plenty of Republican gerrymandering.

No.  According to the official numbers from the Clerk of the House, the totals were:

R: 44,593,666
D: 38,854,459

http://clerk.house.gov/member_info/electionInfo/2010election.pdf (page 59)

[GeorgiaModerate]

https://twitter.com/leahaskarinam/status/996071614763732995

[GeorgiaModerate]

YouGov, May 6-8, 1228 RV

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

[GeorgiaModerate]

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

Cuz the models turned out fantastic for you guys in 2016 right?

If you have an issue with this particular model, do explain why.  I'm sure many people here would be interested.  But pointing to failures of other models in the past is irrelevant.

[GeorgiaModerate]

YouGov, May 13-15, 1231 RV

D 42 (-2), R 37 (+2). 

[GeorgiaModerate]

Is YouGov doing multiple polls per week now?

YouGov has started putting out daily snapshots of Trump approval (not sure if it includes GCB) with 1000-adult samples.  These are included in the 538 database, but I haven't bothered posting them because (a) they're really noisy and (b) it's too much work.  I think they're only doing one larger survey (multiday, 1500 adults) per week, but they may occasionally do something extra.

[GeorgiaModerate]

Y'know, no one is going to learn their damn lesson. We're gonna repeat the doomsaying of the past few weeks within two months or so.

One definition of insanity: doing the same thing over and over again and expecting different results.

[GeorgiaModerate]

That's a very odd way of asking questions...

Can we even count that as a generic ballot poll?

I don't see why not; I'd count it as D 45, R 43.  It's just more nuanced than the typical GCB question, just like some Trump approval polls separate strong approval/disapproval from somewhat approve/disapprove, but others don't.

[GeorgiaModerate]

FWIW, if you switch the filter on the Reuters poll to 5-day rolling average (which is used by the 538 aggregator), it shows R+1.4 (38.1-36.7).  Still surprising, but less so than the R+5.5 shown the other way.

[GeorgiaModerate]

If it's R+6, Republicans don't lose anything but PA-05 and FL-27 and gain MN-1, MN-8, PA-7, both NH seats, NJ-5, FL-7, and PA-8.

>Implying it could be
LMAO

I guess it's time for this periodic reminder:

[GeorgiaModerate]

https://twitter.com/AndrewHClark/status/998948301117493248

Agree with Sean here. Political environment pretty clearly shifting towards Republicans. No, R+6 is not realistic. But could easily see a tied generic ballot in which the Democrats only pick up a few open seats and don't beat an incumbent.

Andrew Clark, worked for Romney and the NRCC. Of course he's creaming over this poll. I want Reuters to release a statement and explain why they screwed up their methodology.

Or it could be an outlier; they happen even with excellent methodolgy.  Let's see what subsequent surveys show.

[GeorgiaModerate]

YouGov, May 20-22, 1500 adults including 1263 registered voters

GCB (asked of RV only): D 43 (+1), R 38 (+1)

Trump approval went from -8 to -13 among RV, and from -11 to -12 among all adults, in this poll.

[GeorgiaModerate]

https://twitter.com/HotlineJosh/status/999314587194417152

I think this tradeoff would make the Democrats quite happy.

[GeorgiaModerate]

https://twitter.com/HotlineJosh/status/999314587194417152

I think this tradeoff would make the Democrats quite happy.

How did educated whites vote in 2006?

How did uneducated whites vote in 2014?


The tradeoff is not good imo. More uneducated whites vote than educated whites in basically every state, besides maybe Massachusetts. There's an even bigger difference in highly competitive Senate seats in places like West Virginia, Indiana, and Nevada.

I interpreted it a little differently: that educated whites were polling strongly D (since 2006 was a D wave) this year, while uneducated whites were polling strongly R.  That would be consistent with other polling this year; and since it also seems that educated voters are highly motivated this year, that would overall be good news for the Democrats.  But your point about the effect in certain states being potentially bad for them is well taken.

[GeorgiaModerate]

Harvard Harris: D 44 - R 37

http://harvardharrispoll.com/wp-content/uploads/2018/05/Final_HHP_May2018_RegisteredVoters_Topline_Memo-Rev1.pdf

Fake news. If you run this through LL's mind model, it comes out to R+1. Politics 101 thx

It's more complex than that!  Let's break down D 44, R 37:

D 44: 4 + 4 = 8.

R 37: 3 + 7 = 10.

10 > 8.  Thus, Republicans are favored.  QED

[GeorgiaModerate]

Remember that last Ipsos poll with a sudden jump to R+5?  Yes, boys and girls, this is what we call an "outlier".  New poll:

D: 42 (+6)
R: 34 (-7)

[GeorgiaModerate]

How many articles and narratives were made over a likely computer error?

Charles Franklin, the Marquette U. pollster, has a long thread about how the partisan balance in Ipsos suddenly changed around May 1, and this has contributed a lot to the increases in Trump's approval and the R position on the generic ballot.  This was written before the new numbers from Ipsos.

https://twitter.com/PollsAndVotes/status/1000008090811445249

[GeorgiaModerate]

This is kind of interesting.  I hadn't realized it's been so long since there was a live interview poll.

https://twitter.com/Nate_Cohn/status/1000809842351656960

[GeorgiaModerate]

Echelon Insights: D+7

Does this pollster have much of a track record?  I don't see them in 538's database.

They have Trump approval at 39/45 among all adults and 39/48 among RV, so that seems plausible.