2018 Congressional Generic Ballot and House Polls Megathread - the original

[GeorgiaModerate]

Ipsos/Reuters, Mar 11-15, 1265 RV

D 45 (+4)
R 35 (-2)

The previous poll may have been an outlier; the two before that were D+10 and D+9.

[GeorgiaModerate]

HOLLY FYCK!!!

How is this a good thing? This just proves the Democratic advantage is packed in already-Dem districts.

An even split in the R-held districts would be very bad news for the GOP.  It would mean that the Democrats would be virtually certain to pick off a bunch of their seats.

[GeorgiaModerate]

Ipsos/Reuters, March 15-19, 1264 RV

D 41 (-1)
R 37 (+1)

[GeorgiaModerate]

Roll Call says average number of House seats in a midterm is 33 for the opposition party.  The polls will tighten.

However, there is an ethics cloud hanging over the Trump administration that voters will go to the polls with and that's a winning hand for Dems. Which translated to victories in PA-18 and AL-Senate

It is highly likely that the Democrats will win way more than just 33 seats.

Take out the adverbs and I'll agree.

[GeorgiaModerate]

I just read through this entire thread trying to find individual district polls.

It's amazing how predictable the responses of some people in here are.

And even after that, you're still here?  I admire your persistence!

[GeorgiaModerate]

The model.

So far this cycle's GCB curve has tracked the model very well, but I encourage everyone not to fall in love with the model and assume that they can't diverge at some point. 

[GeorgiaModerate]

Heads up, a new Fox News poll has the generic ballot at +5 for democrats.

Here: http://www.foxnews.com/politics/interactive/2018/03/25/fox-news-poll-results-325.html

March 18-21, 1014 RV (change from January)

D 46 (+2)
R 41 (+3)

[GeorgiaModerate]

Oh god is this that time when Trump approval goes slightly up and CGB tightens a but and Atlas has a freaks out

Yep, just like every other time.

[GeorgiaModerate]

Fox News Poll 3/25 (Last poll was October 2017)


Democrats - 46 (-4)
Republicans - 41 (+6)

Went from D+15 to D+5.

http://www.foxnews.com/politics/2018/03/25/fox-news-poll-gap-narrows-on-2018-vote-preference.html

There was a poll in January

There was but Fox News has forgotten to mention it, I wonder why? /s

Not only did they leave out the January poll in the trend information, the two they list prior to October (April and June) were 47/42 and 47/41.  January was 44/38 and this month is 46/41.  The 50/35 result in October looks like a clear outlier.

[GeorgiaModerate]

What was in the January poll that they conveniently omitted?

GCB was 44/38.  Here's the whole poll: https://andersonrobbins.com/wp-content/uploads/2018/01/012418_national_january-24-release.pdf

[GeorgiaModerate]

At this point, I have completely given up on tracking the generic ballot polls. All the pollsters, no matter what result they post, acknowledge the fact the GOP is probably going to lose the House in November.

Just like how they said Hillary had a 95% chance of winning in 2016? 😜

I believe 538 had Hillary at about 70% during the week before the election (other forecasters were certainly higher), which meant they gave Trump a 30% chance.  This isn't that unlikely!  Think about a baseball player with a .300 batting average.  How likely is is that he'll make an out in an average at-bat?  That's far from a sure thing, and it's the same likelihood as Hillary's 70% chance.

[GeorgiaModerate]

Ipsos/Reuters, March 22-26, 2244 RV

D 41 (nc)
R 34 (-2)

[GeorgiaModerate]

Morning Consult, March 26-27, 1997 RV

D 41 (-1)
R 37 (-1)

(No net change)

[GeorgiaModerate]

Ipsos/Reuters core political poll (this is different from their daily tracker), March 23-27, 1670 adults (1392 RV)
https://www.ipsos.com/en-us/news-polls/reuters-ipsos-data-core-political-2018-03-28
https://www.slideshare.net/AnneMarieMoran2/reutersipsos-data-core-political-03282018

Among all adults: D 38, R 31 (D+7)

Among RV: D 42, R 35 (D+7)

[GeorgiaModerate]

PPP poll of South Carolina CD 1, which Trump won 54-40 and Romney won 58-40.

https://www.postandcourier.com/cunningham-st-district-poll/pdf_60d4f8e6-3431-11e8-8249-6361aacaeb48.html

Trump approval: 46/47.

Generic Congressional ballot: R 46, D 40.

Mark Sanford favorability: 44/40.

Reelect Sanford or prefer someone new?  Sanford 38, someone new 52.

Prefer a Republican who will support Trump's agenda, or a Democrat who will work across party lines to get results?   D 53, R 40.

The last four viewed together are interesting, to say the least.  I'm not quite sure what to make of them.

[GeorgiaModerate]

At this point, I have completely given up on tracking the generic ballot polls. All the pollsters, no matter what result they post, acknowledge the fact the GOP is probably going to lose the House in November.

Just like how they said Hillary had a 95% chance of winning in 2016? 😜

She did.

Actually, 538 had her at 70% or even less during the final week, meaning Trump was as likely to win as a .300 hitter in baseball was to get a hit -- something that happens all the time.

[GeorgiaModerate]

YouGov, April 1-3, 1244 RV

D 43 (+1)
R 36 (+1)

D+7, no net change.

[GeorgiaModerate]

Morning Consult/Politico, March 29-April 1, 1997 RV

D 41, R 37 (exactly the same as their previous survey).

[GeorgiaModerate]

Harvard Institute of Politics Spring 2018 youth poll, March 8-25, 2631 adults age 18-29 only

D 69 (+4 from Fall 2017 poll)
R 28 (-5)
 
Other items:

Trump approval: 25/72 (no change from Fall).

37% say they will definitely be voting, compared to 23% in 2014 and 31% in 2010.  This includes 51% of young Democrats and 36% of young Republicans.  In 2014, it was 28% of Ds and 31% of Rs; in 2010, it was 35% of Ds and 41% of Rs.

Additional details at http://iop.harvard.edu/spring-2018-poll.

[GeorgiaModerate]

Qpac usually gives solid polls, but this one is seriously f***ed up. 18-29 year olds aren't going to give the GOP even 30% of the vote in November, let alone 42%.

Even good pollsters will occasionally get a wild sample.  Throw it in the average with the others.

[GeorgiaModerate]

YouGov, April 8-10, 1292 RV

D 44 (+1)
R 36 (nc)

The D+8 margin is the highest for YouGov in several weeks.

[GeorgiaModerate]

Muhlenberg College Pennsylvania poll:

Congressional Generic Ballot

Democrats - 47
Republicans - 38

For reference, Democrats lost the 2016 congressional vote 46-54 in 2016.

I would estimate Democrats pick up PA-05, PA-06, PA-07, PA-01, PA-17, with competitive but r-tilting races in PA-10 and PA-16.

Encouraging. Cartwright should be safe in PA-08 with those statewide numbers as well.

FWIW, Muhlenberg has an A rating from 538.

[GeorgiaModerate]

What I think is most telling in these polls is not just the margin in favor of the Democrats but the enthusiasm gap in their favor, as well as the greatly increased motivation among young voters.  If these hold up until November, most of the races that currently look close are going to break Democratic.  It also means that anyone using a likely voter model based on the past two midterms is going to be way, way off.

[GeorgiaModerate]

Haven’t most GCB polls been adults anyways?

No, in most cases they're registered voters.

[GeorgiaModerate]

The advantage with all adults is +10 which makes me curious why their registered voter screen filtered out so many people.

Only 2014 voters, maybe?

I could see that in a likely voter screen (although it would be inaccurate this year).  But a registered voter screen should be as simple as "are you registered to vote? yes/no".