15 February 2016

Polling Wrap-Up, 01-14.02.2016 (except New Hampshire)

So, here is all of the polling for the first two weeks of February, 2016, excepting New Hampshire, for which I put out a special polling round-up HERE, on 09.02.2016, about one week ago. Not quite the blizzard of polls from the end of January, it was still a huge amount: 159 (or more) matchups.

The state polling has brought us once again some interesting state polls. Here again a map of the USA, with the states for which polled were released between 01-01.02.2016:



All of the states you see colored in green are states where polls were released in the first two weeks of February. All of them except one were polled for either the GOP nomination, the DEM nomination, or both. Virginia was polled only for D-vs-R presidential GE matchups.

And away we go:

GOP NOMINATION: NATIONAL POLLING

Baseline from the end of January: Trump +20.12

Morning Consult (R), released 01.02.2016: Trump +29
NBC News, Survey Monkey, released 02.02.2016: Trump +19
PPP (D), released 04.02.2016: Trump +4
IPSOS/Reuters, released 04.02.2016Trump +22 (RRV + IRV, R-leaning), Trump +22 (RRV only), Trump +24 (IRV, R-leaning only)
Quinnipiac, released 05.02.2016: Trump +9
Week 1 margin: Trump +16.6

Rasmussen, released 08.02.2016: Trump +10
Morning Consult, released 09.02.2016: Trump +19 (big drop from the Morning Consult from the week before)
NBC / News Monkey, released 09.11.2016: Trump +15
IPSOS/Reuters, released 10.02.2016Trump +13 (RRV + IRV, R-leaning), Trump +13 (RRV only), Trump +16 (IRV, R-leaning only)
Morning Consult (R), released 12.02.2016: Trump +27
Week 2 Margin (w/o repeaters):  Trump +16.25

That makes for 10 national polls in 2 weeks, four of them being repeaters.  Trump's weekly aggregate moved down from +20.1 at the end of January to +16.6 in the first week of February and to +16.3 in the second week.  The PPP (D) poll, in very stark contrast to the 2 polls before it, looks like a mathematical outlier, but the Quinnipiac poll that followed also shows Trump in a single digit lead. In all but one poll (Rasmussen), Texas Senator Ted Cruz is in second place.

So, although we see the race tightening statistically somewhat in national polling, Donald Trump is still very much in the lead across the board.


GOP NOMINATION: STATE POLLING

Alabama:
Overtime Politics, released 04.02.2016: Trump +16

Arkansas:
Hendrix College / Talk Business, released 07.02.2016: Cruz +4 (3-way tie)

Florida:
Florida Southern College Center (FSCS), released 09.02.2016: Trump +7 (Rubio in 2nd)

Georgia:
Landmark Communications (R), Rosetta Stone / WSB TV-2/4, released 05.02.2016: Trump +9

Iowa:
Quinnipiac, released 01.02.2016 (on the day of the IA caucuses): Trump +7
(This poll came in just before the caucuses actually began that day)

Massachusetts:
Overtime Politics, released 12.02.2016: Trump +31

Michigan:
"Target Insyght" / MRS / IMP, released 05.02.2016: Trump +14
Mitchell Research (R) / Fox 2, released 08.02.2016: Trump +21
Average: Trump +17.5

New York:
Siena, released 08.02.2016: Trump +18

North Carolina:

High Point University, released 08.02.2016 Trump +4 (statistical 2-to-3 way tie)

Oklahoma:
Sooner Poll, released 09.02.2016: Trump +5

South Carolina:

Opinion Savvy / Augusta Chronicle / FOX 5, released 12.02.2016: Trump +16.7
CBS News / YouGov, released 14.02.2016: Trump +22
ARG, released 12.02.2016: Trump +20
Average: Trump +19.6

That makes for 14 polls from 11 states, plus the 29 polls from New Hampshire, already recorded HERE, for a  total of 43 GOP state polls from 12 states. Trump has won 41 of those 43 and two of them are statistical ties. The huge margin out of Massachusetts (Trump +31) is worth noting, also the lean margin for him out of Oklahoma (The Sooner Poll is the "gold standard" for that state).

The long and short of it is: so long as there are many anti-Trumps in the mixt, he is going to sail easily to the nomination.  However, the latest ARG shows Kasich in 2nd. ARG was also the pollster who showed Kasich in 2nd in New Hampshire.  If Bush drops out and his support goes to Kasich, who may continue picking up delegates, and Cruz will certainly be picking up delegates, this could become a 3-4 man race, at least for a while. After Rubio crashed in New Hampshire, his important testing ground is South Carolina, where the demographic cards are stacked against him.

GOP grand total: 43 state polls + 10 national polls = 53 matchups (at least).

All of the toplines, survey sizes, MoEs and some important internals and crosstabs for the GOP polling can be found HERE.


-----------------------------------------------------------------------------------


DEM NOMINATION: NATIONAL POLLING

Baseline from the end of January, 2016: Clinton +15.8

Morning Consult (R), released 01.02.2016: Clinton +16
NBC News, Survey Monkey, released 02.02.2016: Clinton +11
PPP (D), released 04.02.2016: Clinton +21
IPSOS/Reuters, released 04.02.2016Clinton +11 (DRV +IRV, D-leaning), Clinton +15 (DRV only), Sanders +5 (IRV, D-leaning only)
Quinnipiac, released 05.02.2016: Clinton +2
Rasmussen, released 05.02.2016: Clinton +18
Week 1 average: Clinton +13.8

Morning Consult, released 09.02.2016: Clinton +13
NBC / News Monkey, released 09.11.2016: Clinton +12
IPSOS/Reuters, released 10.02.2016Clinton +2 (DRV +IRV, D-leaning), Clinton +12 (DRV only), Sanders +29 (IRV, D-leaning only)
Morning Consult (R), released 12.02.2016: Clinton +7
Week 2 average: Clinton +10.3


That makes for 10 national polls over the last two week, of which Clinton has won all 10. We see two interesting parallels to the GOP national polling: first, we see a number that looks like an outlier. For the GOP, it was PPP (D). For the DEMS, it is Quinnipiac. We also see very varied numbers from the three Morning Consult polls, just as we saw on the GOP side.

There is no doubt that her margin over Vermont Sen. Bernie Sanders is narrowing, or maybe not. Among registered and likely Democrats, she is still way on top. Among Independents who are likely to vote in the D primaries in states that have semi-open primaries/caucuses, Sanders dominates.

But no matter how you slice it, Clinton is still in the driver's seat.



DEM NOMINATION: STATE POLLING

Alabama:
Overtime Politics, released 04.02.2016: Clinton +11

Arkansas:
Hendrix College / Talk Business, released 07.02.2016: Clinton +32

Florida:
Florida Southern College Center (FSCS), released 09.02.2016: Clinton +17

Georgia:
Landmark Communications (R), Rosetta Stone / WSB TV-2/4, released 05.02.2016: Clinton +41.8

Iowa:
Quinnipiac, released 01.02.2016 (on the day of the Iowa Caucuses): Sanders +3
(was released shortly before the caucuses began that day)

Maine:
Overtime Politics, released 06.02.2016: Sanders +15

Massachusetts:
Overtime Politics, released 12.02.2016: Clinton +6

Michigan:
"Target Insyght" / MRS / IMP, released 05.02.2016: Clinton +32
Mitchell Research (R) / Fox 2, released 08.02.2016: Clinton +29
Average: Clinton +25.5

Nevada:

TargetPoint / Washington Free Beacon, released 12.02.2016: -TIE- (45/45)
analysis below

New York:

Siena, released 08.02.2016: Clinton +21

North Carolina:

High Point University, released 08.02.2016: Clinton +26

Oklahoma:

Sooner Poll, released 09.02.2016: Clinton +16

South Carolina:

CBS News / YouGov, released 14.02.2016: Clinton +19
ARG, released 14.02.2016: Clinton +38
Average: Clinton +28.5


That makes for 15 polls from 13 states, plus the 26 DEM polls from New Hampshire, which you can see HERE. Total: 41 polls from 14 states on the Democratic side.

Excluding Iowa, as the poll literally came in just hours before the caucuses began, Sanders is leading in one state: Maine (according to Overtime Politics, a scenario I see as very plausible). It is also a lean margin for Clinton in Massachusetts, also understandable, considering the hard-Left tilt of the state over all. But everywhere else, save Nevada (at least for now), Clinton is dominating.  Especially noteworthy are the margins in Michigan and both Carolinas. In South Carolina, it is actually quite impossible to have two margins from polls released the very same day, where one margin is double the size of the other margin and both of them are massive landslide margins.  In most all polling, Clinton has been at about +40 here and I suspect that her landslide in SC will be larger than Sanders' was in New Hampshire.

Now, to this quirky little poll from Nevada, here is some food for thought:

TargetPoint is a Republican polling firm and many of the extra questions they asked in the poll come dangerously close to being "push-poll" questions. Also, the Washington Free Beacon is a hard Right leaning news publication. In fact, it is so hyperpartisan that you can find this on the permanent front page right now:


So, why a so unbelievably hard-Right publication would want to sign up with a Republican pollster to test a Democratic race is somewhat of a mystery to me.  But THIS data point tells me how off kilter the poll is:



Considerably more than just 17% of NV Democrats belong to a Union. In fact, the Culinary Union vote from Las Vegas will be the deciding factor in the NV primary, and that majority of the Culinary Union workers from Las Vegas are Latinos, who are swinging extraordinarly heavily for Clinton.

Also, the poll lied about "likely Democrats". Only 67% of the poll respondents are definitely going to caucus, and only 72% of all of the survey respondents are Democrats.


I would really like to know exactly WHO paid for this poll. And I would put good money on the table that it was either the Trump or the Cruz campaign. Anyone wanna bet?

Facit: Sanders is making a dent in Clinton's margins in some places and he may actually have an advantage in some caucus states, but Clinton is absolutely dominating in ALL of the South, on the Atlantic seaboard and at least in one critical Midwest state.  My feeling is still that former Secretary of State Clinton's path to the nomination is lighter than Trump's path to the GOP nomination.

Grand total: 51 DEM matchups, national and state polling combined. Running subtotal, D + R together: 104 matchups (at least).

All of the toplines, survey sizes, MoEs and some important internals and crosstabs for the DEM polling can be found HERE.

----------------------------------------------------------------------------------------------

D vs. R matchups: NATIONAL POLLING

Morning Consult (R), released 01.02.2016:
waiting on PDF, but this link claims that both DEMS easily lead the GOP field.
Absent newer statistics, there is not much for me to say about this one. However, two more MC polls came out afterward.

Quinnipiac, released 05.02.2016: 8 matchups, 3 of them Clinton vs. GOP, 3 of the Sanders vs. GOP, 2 of them Sanders vs. GOP vs. Bloomberg. 4 D wins, 2 R wins, 2 ties. The poll claims absolute parity in voter party affiliation: D32/D32. Ain't gonna happen...

Morning Consult, released 09.02.2016: 9 matchups, 5 of them Clinton vs. GOP, 4 of them 3 way matchups, either Clinton or Sanders vs. GOP vs. Bloomber. 8 D wins, 1 GOP win. The survey reflects a D+5 party affiliation electorate, much closer to the actual stats from 2008 and 2012.

IPSOS/Reuters, released 10.02.2016: 8 matchups, 2 Clinton vs. GOP, 2 Clinton vs. GOP vs. Bloomberg, same scenario for Sanders. 8 D wins. The survey reflects a D+9 party affiliation electorate, is too far off to the Left in this case.

Morning Consult (R), released 12.02.2016: 4 matchups, all 3-ways: 2 Clinton vs. GOP vs. Bloomberg, same scenario for Sanders. 4 D wins.

That makes for 29 national matchups that I can verify in the last 14 days: 24 D wins, 2 ties, 3 R wins.


D vs. R matchups: STATE POLLING


Florida:
Florida Southern College Center (FSCS), released 09.02.2016: 8 matchups, 4 Clinton vs. GOP, 4 Sanders vs. GOP. 6 D wins, 2 GOP wins. The party affiliation screen is WAY, WAY, WAY off to the right in the poll. See: HERE.

New York:
Siena, released 08.02.2016: 12 matchups, 6 Clinton vs. GOP, 6 Sanders vs. GOP. 12 D landslide wins, and yet, the +17 over Rubio is disturbing. The last time a Democrat won NY State by "only" +18 was John Kerry, in 2004, and he lost to Pres. Bush (43).  The good news is that NY margins in polling tend to be smaller than they actually end up being on election night. Sanders beats Trump by +33 here. Were this to hold, it would be the largest D margin in this state since LBJ, 1964.

The most ominous warning sign for the GOP:

Virginia:
Roanoke College, released 03.02.2016:  6 matchups, 3 Clinton vs. GOP, 3 Sanders vs. GOP. The Ds win all 6 easily. Clinton and Sanders lead Trump by +15 and +22, respectively.  This is a sure sign thast this state is being cemented into the so-called "Blue Wall", unless the GOP does something about it. The only margin that is under Obama's +3.87% win in the Old Dominion from 2012 is Clinton's +3 over Rubio. Once again, in terms of actual state-to-state polling, Rubio comes out as the most viable opponent to Clinton.


This makes for 26 state matchups, 14 of them from two critical battleground states and makes for 55 national and state D-vs-R matchups from the first two weeks in February. This makes for a grand total of 159 (at least) matchups for this polling wrap-up.

Now, it is still a long time until November, but Obama never had these kinds of numbers in Virginia like what we see against Trump, Bush and Cruz.


All of the toplines, survey sizes, MoEs and some important internals and crosstabs for the D-vs-R polling can be found HERE.






No comments:

Post a Comment

Constructive comments and critique are always welcome. Please keep it polite and respectful.