02 May 2015

Hillary polling wrap-up, last 10 days of April, 2015

From 22 April 2015 to 01 May, 2015, there was national polling, Hillary Clinton (D) vs. GOP field. There was also polling for the same in these states:








National:


Quinnipiac, taken from 16-21 April, 2015, released 23 April 2015. 1,353 RV, MoE= +/-2.7
(Values in parentheses are from the previous Quinnipiac, released 05 March 2015) 

Clinton 45 (46) / Rubio 42 (41), margin = Clinton +3 (+5)
Clinton 46 (47) / Paul 42 (41), margin = Clinton +4 (+6)
Clinton 45 (46) / Christie 40 (39), margin = Clinton +5 (+7)
Clinton 47 (47) / Huckabee 42 (40), margin = Clinton +5 (+7)
Clinton 46 (48) / Walker 40 (44), margin = Clinton +6 (+4)
Clinton 48 (48) / Cruz 41 (38), margin = Clinton +7 (+10)
Clinton 46 (45) / Bush, J. 39 (42), margin = Clinton +7 (+3)


This means that the Quinnipiac margins for Hillary, who wins all 7 match-ups, are from a lean +3 to a solid +7. In 6 of seven matchups, the margin over the previous Qpiac poll shifted exactly two points. In the case of Ted Cruz, it shifted 3 points. 2 points is well within the MoE of both polls, which means that the margin from the last poll could easily have also been this margin, or the margin from this poll could have easily also been the previous margin. In 5 of seven matchups, the margin has tightened, but in two matchups (Bush, J and Walker), it has widened.

Two polls from YouGov / The Economist, and they cover different candidates from the GOP:

1.) YouGov / The Economist, taken from 18 to 20 April, 2015, released 23 April 2015, 1,000 adults, MoE = +/-4.1

2.) YouGov / The Economist, taken from 25 to 27 April, 2015, released 01 May 2015, 1,000 adults, MoE = +/-4.1

In the first poll, it's Clinton vs. Paul, Cruz and Rubio
In the second poll, it's Clinton vs.Christie, Bush and Huckabee.

Together, it's:

Clinton 46  / Rubio 42 , margin = Clinton +4
Clinton 46  / Paul 40 , margin = Clinton +6
Clinton 46  / Cruz 37, margin = Clinton +9
Clinton 46  / Christie 34 , margin = Clinton +12
Clinton 46  / Bush, J. 34 , margin = Clinton +12
Clinton 48  / Huckabee 34, margin = Clinton +14

Interesting to note is that the margins from the earlier poll (Rubio, Paul, Cruz) are all leaner than the margins for the later poll (Christie, Bush J and Huckabee). If you look at both .pdfs, you wil see that the category of "will not vote" is between 5-7% for the first poll, but between 6-9% in the later poll.  YouGov is a firm out of Great Britain that does Internet only polls. I did an analysis of YouGov's performance in the last presidential election (2012), you will find the results at the link, but the long and short of it is that, of 27 end-polls (YouGov put out more end polls than any other pollster in 2012), it ended up with a mathematical bias of 1.63 points to the Right for all 27 polls and 1.75 points to the Right for the 12 battlegrounds, making it one of the better pollsters in 2012. YouGov absolutely nailed the margins in Ohio, North Carolina and Georgia.


FOX NEWS POLL: taken from 19 to 21 April, 2015, released on 23 April, 2015, 1,012 RV, MoE = +/-3.0

(Values in parentheses are from the previous FOX NEWS POLL, released 29 January 2015) 

Clinton 46 (47) / Paul 43 (44), margin = Clinton +3 (+3)
Clinton 46 (48) / Rubio 42 (41), margin = Clinton +4 (+7)
Clinton 45 (48) / Bush, J. 41 (43), margin = Clinton +4 (+5)
Clinton 45 (48) / Christie 40 (42), margin = Clinton +5 (+6)
Clinton 47  / Cruz 42 , margin = Clinton +5 
Clinton 47  / Huckabee 42 , margin = Clinton +5 
Clinton 46  / Walker 40 , margin = Clinton +6 

In the FOX NEWS poll, Hillary wins all 7 matchups by between +3 and +6. Many of a margins are very, very close to the Quinnipiac margins. The +5 margin against Ted Cruz is the leanest margin recorded to date, he is usually between 7 and 20 points behind Clinton nationally.

So, between these three pollsters, that makes for 20 national matchups, all of which Clinton has won.



North Carolina:

A state that was a bitter battleground in both 2008 and 2012, which is also a break from the electoral history of this state, polling shows the Tarheel State heading toward battleground status again in 2016.

Elon University poll: taken from 20-24 April, 2015, released 27 April, 2015, 677 RV, MoE = +/-3.77.

Clinton 47 / Bush, J. 44 , margin = Clinton +3


SUSA (Survey USA): taken from 22-27  April, 2015, released 27 April 2015, 1,070 LV, MoE = +/-3.1

Clinton 45  / Bush, J. 43 , margin = Clinton +2
Clinton 47  / Paul 43 , margin = Clinton +4
Clinton 46  / Rubio 41 , margin = Clinton +5
Clinton 47  / Cruz 42, margin = Clinton +5
Clinton 48  / Walker 39, margin = Clinton +9

Clinton 44  / GENERIC 42 , margin = Clinton +2


SUSA put out 5 matchups of North Carolina, showing Clinton leading the GOP field by between +2 and +9. This is also the first state poll in a long time to also list a "generic" matchup, which I find totally worthless at this point in time. The Hillary vs. Bush matchups from both SUSA and Elon are in line with each other. The SUSA website puts out their results in a fantastic way, I can recommend that you go to the above link, for the demographic breakdowns are right there to see quite easily.

SUSA was one of the very best pollsters of 2008, also of the 2010 mid-terms. It polled less in 2012, but nailed the critical battleground state of Ohio. In 2014, the SUSA polling in Kentucky was, however, way off. That being said, SUSA is a truly independent polling firm that has no dog in the race at all and is willing to take contracts from pretty much anyone and for this reason, has no interest in peppering the numbers in any way at all.

Another important note about the SUSA poll: the female vote.

First, take a look at the North Carolina exit poll from 2012

Obama 51 (55) / Romney 49 (44), margin = Obama +2 (+11)

Remember that +2 for Obama in 2012. Now, go look at the margins for Clinton in the female vote in the SUSA Poll: +11 over Bush, +12 over Cruz and Paul, +15 over Rubio, +16 over Walker, +9 over "generic".  If those kind of margins in the female vote hold for Clinton until the election, then the GOP is going to have a very, very hard time retaining this state in the GOP column.

North Carolina is also the home state of PPP (D) polling, which usually puts out an NC poll at the end of each month, but did not in April.



Iowa:




PPP (D): taken from 23-26 April, 2015, released on 30.04.2015, 1,219 RV, MoE = +/-2.8

(Values in parentheses are from the previous PPP (D) poll, released 30 September 2014, exactly 7 months ago) 


Clinton 46  / Rubio 44, margin = Clinton +2
Clinton 46 (46)  / Huckabee 44 (42) , margin = Clinton +2 (+4)
Clinton 45 (47) / Paul 43 (42), margin = Clinton +2 (+5)
Clinton 45 (43) Bush, J. 42 (43), margin = Clinton +3 (+/-0)
Clinton 45 (45) / Christie 40 (39), margin = Clinton +5 (+6)
Clinton 48  / Walker 41 , margin = Clinton +7 
Clinton 48  / Carson 41 , margin = Clinton +7 
Clinton 48  / Perry 41 , margin = Clinton +7 
Clinton 49 (47)  / Cruz 42 (39) , margin = Clinton +7 (+9) 


In the PPP (D), of which we can compare 5 of 9 matchups to a poll from 7 months ago, Clinton wins all nine matchups, but four of them are close. Iowa has shown very close numbers in most of the 23 polls that have occured this far, making for 109 matchups, of which Clinton has won 96, the GOP has won 6 and there were 7 absolute ties.




New Hampshire:

Townhall/Gravis (R) poll: taken 21-21 April, 2015, released 27 April 2015, 1,117 RV, MoE = not listed

(Values in parentheses are from the previous Townhall/Gravis (R) poll, released 21 March, 2015)


Walker 47 (43)  / Clinton 44 (46) , margin = Walker +3 (Clinton +3)
Paul 45 (42)  / Clinton 44 (45) , margin = Paul +1 (Clinton +3)
Clinton 42 (43)  / Bush, J. 42 (40) , margin = TIED (Clinton +3)
Clinton 45  / Rubio 44 , margin = Clinton +1
Clinton 43  / Christie 41 , margin = Clinton +2
Clinton 47  / Cruz 44, margin = Clinton +3

The Townhall/Gravis (R) poll gives us a chance to compare three matchups to the poll from the month before (Clinton vs. Walker, Paul and Bush) and in all three matchups, according to Gravis, the GOP has made gains. The candidate who has made the most gains is Walker, who went from -3 to +3 against Hillary, a 6 point shift, the largest shift of any of these polls. The spectrum ranges from Walker +3 to Hillary +3, one matchup up is a mathematical absolute tie (although with such unbelievably high name recognition as both the Clintons and the Bushes have, I find it hard to believe that both candidates are mired at 42%). In essence, Townhall / Gravis paints a picture of a very tied up race. This is of course very possible, but the numbers are so very different from the PPP (D) poll that was taken 10 days before. 

In 2012, the end polling from the 7 states from Gravis (R) produced a mathematical bias of +2.86 to the Right, making Gravis one of the three worst pollsters of 2012, alongside Rasmussen (+4.50 to the Right) and Pulse (a Subsidiary of Rasmussen, +2.60 to the right). Also, the end poll of New Hampshire by Gravis showed Obama +1, he won by almost +6, so the poll was off to the Right in this state by 5 full points, which makes for a bad track record.

I am not writing this to paint Gravis (R) in a bad light, but that is indeed it's polling record.  In fact, to give a really good idea of this, here is a screenshot of the end-statistics from the end-polling in 2012:









Virginia:


Christopher Newport poll, taken from 13-14 April, 2015, released 27 April.2015, 658 EV, MoE = +/-4.6

(Values in parentheses are from the previous Christopher Newport poll, released 12 Feb, 2015)


Bush 48 (43)  / Clinton 46 (48) , margin = Bush +2 (Clinton +5)
Clinton 49  (44) / Paul 47 (42) , margin = Clinton +2 (+2)
Clinton 47  (44) / Christie 45 (39) , margin = Clinton +2 (+5)
Clinton 49  (44) / Huckabee 46 (41) , margin = Clinton +3 (+3)
Clinton 49  (51) / Rubio 45 (42) , margin = Clinton +4 (+9)
Clinton 48  (45) / Walker 43 (40) , margin = Clinton +5 (+5)
Clinton 49  / Cruz 44, margin = Clinton +5

6 of 7 matchups can be compared to the previous poll. In all cases, Clinton has lost ground and for the first time in a Virginia poll, Bush (Jeb) has taken a lead. But he is the only one who leads. Just to illustrate, there have now been 19 polls from the Old Dominion, with 70 matchups, and Clinton has won 66 matchups, the GOP has won 2 (Bush, Jeb being one of those two right now) and there have been two absolute ties. According to this poll, the margins range from Bush +2 to Clinton +5, definitely showing battleground possibilities for this state.




California:

An older poll that slipped through pretty much everyone's fingers came through, it is a quirky poll of California:

Emerson College Poll, taken from 2-8 April, 2015, released 13 April, 2015, 882 RV, MoE = +/-3.2


Clinton 53  / Bush, J. 47 , margin = Clinton +6
Clinton 53  / Walker 47, margin = Clinton +6
Clinton 55  / Cruz 44, margin = Clinton +11

Those are very, very lean Democratic margins for California standards. In fact, the Emerson Poll reports says:


"A new poll conducted in California by the Emerson College Polling Society (ECPS) suggests that the perennially blue state, which has not been won by a Republican presidential candidate since 1988, may be in play for the 2016 election."

Obama won Californa by +23 in 2012 and by +24 in 2008. The end-polling aggregate for California in 2012 predicted: Obama +15.80, so the end-aggregate was off by 7 points to the Right.

I am not saying it's impossible that the margins are this small. It's possible, but it's very unlikely. Also, Emerson College, from Massachusetts, usually does polling on the East Coast and it very much out of it's area, so say the least.

I will also remind that in 2008, there was one Siena Poll from New York that showed Obama up on McCain by only +4 in April of that year. In November, he won by almost +26.

But I will surely be keeping an eye on California to see if a pattern emerges.

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


In conclusion, in this report, I had analyzed 52 matchups from 9 pollsters, Hillary vs. GOP field.  Of those 52 matchups, Hillary "won" 49 matchups, the GOP "won" 2 matchups and there was one absolute tie.

But many of those wins are technically, statistically also ties, for they are well within the MoE.

Speaking of MoE, at +/-4.6, the MoE for the CNU poll of Virginia is way too high, and Gravis (R)did not even release it's MoE. Not only that, Gravis (R) does not release it's internals. Most polling firms do.

The pattern we are seeing right now, with candidates actually entering the race for real, is that the numbers are tightening up some. Against Hillary, Marco Rubio has gained the most ground, but nationally she is still undoubtedly leading the GOP field.

What I hope to see next month, after Fiorina, Carson and Huckabee have entered the race, would be some fresh polling from Ohio, Pennsylvania, Virginia, Florida and North Carolina. Those five states together, the "Quintefecta", worth 95 electoral votes in 2016, will decide the race.

Sum-totals, to date. 

314 polls since 2013, 1,238 matchups (national and state matchups - from a mixture of RED and BLUE states), of which Hillary has won 994 (80.29%), the GOP has won 212 (17.12%) and there have been 32 absolute ties (2.58%). This statistic of a 4 to 1 ratio has barely budged over 2 years now.


The next report comes on May 17th, 2015. And the exhaustive Hillary vs. GOP report no. VIII comes out in June.


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















No comments:

Post a Comment

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