2018 Congressional Generic Ballot and House Polls Megathread - the original

[Dr. Arch]

Another fun graph:

....oh my god

sidenote: what happened in 1998??

Bill Clinton, Impeachment of

Ah, of course.

[Blackacre]

Another fun graph:

....oh my god

sidenote: what happened in 1998??

Bill Clinton, Impeachment of

Ah, of course.

Im also noticing that, outlier nonwithstanding, this shows that special elections undershoot Dem performances (1988, 1996, 2002, 2006, 2008, 2012) more than it overshoots Dems (2000, 2014). It's even undershot Dems by huge amounts before. If the difference observed in 2006 is observed again in 2018, that's a D+17 election.

[Gustaf]

Another fun graph:

....oh my god

sidenote: what happened in 1998??

Bill Clinton, Impeachment of

Ah, of course.

Im also noticing that, outlier nonwithstanding, this shows that special elections undershoot Dem performances (1988, 1996, 2002, 2006, 2008, 2012) more than it overshoots Dems (2000, 2014). It's even undershot Dems by huge amounts before. If the difference observed in 2006 is observed again in 2018, that's a D+17 election.

I'd guess that is due to lower Democratic turnout and given the enthusiasm gap that seems less likely to hold true this year.

[Mr. Morden]

Does the "special elections index" do anything to control for which party is the incumbent in the special election?  The thing with special elections is that, by definition, whoever is the incumbent in that party isn't running, and so has no incumbency advantage.  Whereas in the midterms as a whole, there are plenty of seats up where the incumbent is running....but if the special elections are held in both Dem- and GOP-held districts, then I guess this effect cancels out.

In 2017, almost all (all?) of the special elections were held in Republican-held seats.  Was that the case in past years as well, that the incumbent party was the one holding all the seats that were up in special elections?  If not, then I'm not sure if this past year's results are going to be so predictive.

[Blackacre]

Does the "special elections index" do anything to control for which party is the incumbent in the special election?  The thing with special elections is that, by definition, whoever is the incumbent in that party isn't running, and so has no incumbency advantage.  Whereas in the midterms as a whole, there are plenty of seats up where the incumbent is running....but if the special elections are held in both Dem- and GOP-held districts, then I guess this effect cancels out.

In 2017, almost all (all?) of the special elections were held in Republican-held seats.  Was that the case in past years as well, that the incumbent party was the one holding all the seats that were up in special elections?  If not, then I'm not sure if this past year's results are going to be so predictive.

That usually is the case, because the incumbent party's numbers are inflated by House members going to the cabinet, for instance

[Mr. Morden]

Does the "special elections index" do anything to control for which party is the incumbent in the special election?  The thing with special elections is that, by definition, whoever is the incumbent in that party isn't running, and so has no incumbency advantage.  Whereas in the midterms as a whole, there are plenty of seats up where the incumbent is running....but if the special elections are held in both Dem- and GOP-held districts, then I guess this effect cancels out.

In 2017, almost all (all?) of the special elections were held in Republican-held seats.  Was that the case in past years as well, that the incumbent party was the one holding all the seats that were up in special elections?  If not, then I'm not sure if this past year's results are going to be so predictive.

That usually is the case, because the incumbent party's numbers are inflated by House members going to the cabinet, for instance

That would only be relevant in the first year of a presidency though (there usually aren't so many vacancies to fill when a president's in his 3rd, 5th, or 7th year in office), and the graph shown above is for every single congressional election, not just midterms after a president's first two years in office.  And even with that, I think Trump made more appointments from the House than most other presidents.  According to this, for example:

https://en.wikipedia.org/wiki/List_of_special_elections_to_the_United_States_House_of_Representatives

George W. Bush made only one appointment of a sitting member of the House in 2001, and so just 4 out of 7 of the special elections that year were Republican-held seats.  Whereas 5 out of the 6 special elections in 2017 were Republican-held seats.

[junior chįmp]

Now that the GOP's bump has receeded, back to our 🌊 tsunami

New Marist Poll - Generic Congressional Ballot D's Lead By 11

[GeorgiaModerate]

Now that the GOP's bump has receeded, back to our 🌊 tsunami

New Marist Poll - Generic Congressional Ballot D's Lead By 11

That's 49/38 D.  Their previous poll (last month) was 46/40 D.

[KingSweden]

DOOOOOOOOOM

In all seriousness that’s a good poll, but it’s only one poll; and I’m not getting excited unless I see corroboration

[Gass3268]

DOOOOOOOOOM

In all seriousness that’s a good poll, but it’s only one poll; and I’m not getting excited unless I see corroboration

G. Elliot Morris' graph of historical data suggests we should be nearing the high point for the White House Party.

[Gass3268]

Also don't look now, but the average is up to ~7.5 D, up about 2.5 points from earlier in the week.

[Obama-Biden Democrat]

Now that the GOP's bump has receeded, back to our 🌊 tsunami

New Marist Poll - Generic Congressional Ballot D's Lead By 11

Beautiful!

[Landslide Lyndon]

DOOOOOOOOOM

In all seriousness that’s a good poll, but it’s only one poll; and I’m not getting excited unless I see corroboration

G. Elliot Morris' graph of historical data suggests we should be nearing the high point for the White House Party.

Nah, that will be probably in September after Labour Day.

[Gass3268]

DOOOOOOOOOM

In all seriousness that’s a good poll, but it’s only one poll; and I’m not getting excited unless I see corroboration

G. Elliot Morris' graph of historical data suggests we should be nearing the high point for the White House Party.

Nah, that will be probably in September after Labour Day.

Based on the graph it appears they get a bump there, but not to the extent of the post-Holidays/SOTU bump.

[Blackacre]

Yeah. The bump is smaller and so is their absolute position on the GCB

[GeorgiaModerate]

Ipsos/Reuters, Feb 4-8, 2141 RV

D 41 (-2)
R 38 (+3)

[UncleSam]

I wonder what the reason for this downturn is

Just Rs coming home or a reaction to the budget bill discussions / immigration talks?

[Amenhotep Bakari-Sellers]

Dems can win anywhere between 11 and 42 seats in the House.  This is a violatile electorate, the last 3 Presidents they have lost 1 or both houses of Congress.

[LimoLiberal]

NEW POLL: PA-15 (current version)

http://www.mcall.com/news/nationworld/pennsylvania/capitol-ideas/mc-nws-15th-district-congress-poll-dent-trump-casey-20180208-story,amp.html?__twitter_impression=true

Remember this is a Trump +8 district

Generic D: 38
Generic R: 35

• Casey above water 34-31
  Wolf above water 37-32
  Trump disapproval 56%, approval not given.
  Tax bill above water 42-39
  Dent highly popular.

[Holy Unifying Centrist]

NEW POLL: PA-15 (current version)

http://www.mcall.com/news/nationworld/pennsylvania/capitol-ideas/mc-nws-15th-district-congress-poll-dent-trump-casey-20180208-story,amp.html?__twitter_impression=true

Remember this is a Trump +8 district

Generic D: 38
Generic R: 35

• Casey above water 34-31
  Wolf above water 37-32
  Trump disapproval 56%, approval not given.
  Tax bill above water 42-39
  Dent highly popular.

The voters in this district can't make up their mind on anything besides Trump apparently.

Also it says Trump's approval in the district is 40%.

[Holmes]

Cool cool. But the PA-15 in the redistricting the legislature made is very different, and while hat map probably won't be approved, it's probably not a stretch to believe it will end up have very different lines.

[Gass3268]

NEW POLL: PA-15 (current version)

http://www.mcall.com/news/nationworld/pennsylvania/capitol-ideas/mc-nws-15th-district-congress-poll-dent-trump-casey-20180208-story,amp.html?__twitter_impression=true

Remember this is a Trump +8 district

Generic D: 38
Generic R: 35

• Casey above water 34-31
  Wolf above water 37-32
  Trump disapproval 56%, approval not given.
  Tax bill above water 42-39
  Dent highly popular.

The voters in this district can't make up their mind on anything besides Trump apparently.

Also it says Trump's approval in the district is 40%.

With this poll, it is pretty safe to say that a compact Leigh Valley district would go Democratic.

[Brittain33]

This is fascinating. The Upshot at the NYT estimates that the impact of court-ordered redistricting in Pa., Va., Florida, and NC, Republican retirements, and Democratic recruitments has added the equivalent of 2 points to the Democratic advantage on the generic ballot. In other words, it's made it easier for Democrats to take control equivalent to 2 points on the generic ballot but in a way that won't show up in the polls.

https://www.nytimes.com/2018/02/12/upshot/big-republican-advantages-are-eroding-in-the-race-for-house-control.html

[Young Conservative]

This is fascinating. The Upshot at the NYT estimates that the impact of court-ordered redistricting in Pa., Va., Florida, and NC, Republican retirements, and Democratic recruitments has added the equivalent of 2 points to the Democratic advantage on the generic ballot. In other words, it's made it easier for Democrats to take control equivalent to 2 points on the generic ballot but in a way that won't show up in the polls.

https://www.nytimes.com/2018/02/12/upshot/big-republican-advantages-are-eroding-in-the-race-for-house-control.html

Funny how the gerrymandering of California and Maryland is ok but NC and VA isn’t.

[Gass3268]

This is fascinating. The Upshot at the NYT estimates that the impact of court-ordered redistricting in Pa., Va., Florida, and NC, Republican retirements, and Democratic recruitments has added the equivalent of 2 points to the Democratic advantage on the generic ballot. In other words, it's made it easier for Democrats to take control equivalent to 2 points on the generic ballot but in a way that won't show up in the polls.

https://www.nytimes.com/2018/02/12/upshot/big-republican-advantages-are-eroding-in-the-race-for-house-control.html

Funny how the gerrymandering of California and Maryland is ok but NC and VA isn’t.

California wasn’t gerrymandered.