Polling Flood! Eight Polls Point to... a Stable Race-NH,SC,NJ and National -Today in Primary Polling

THE STATES

New Hampshire

CNN/WMUR/UNH (Fivethirtyeight rating: UNH B, WMUR ungraded, CNN graded with a different partner)

Clinton 40

Sanders 50

O’Malley 2

Sanders’s 10 point advantage here is stronger than what we’ve seen in most recent polling, but it also represents an improvement for Clinton over their last poll, where Sanders led by 16 (and Biden was included at 14%). Since then, Clinton has gained 10 points, Sanders has gained 4, and O’Malley has dropped 1.

This poll was also slightly better for Sanders than other polling back in September, when they last released, as well. Overall, Biden’s exit from the race appears to have helped Clinton, but Sanders maintains a solid lead in this poll. 

One good sign for Clinton- she has a solid seven point lead among Democrats, while Sanders’s strength comes from his lead among New Hampshire’s “undeclared” voters who can choose either primary. These voters are already more interested in the GOP race, and if too many decide that the GOP race is a better place to exercise their vote Sanders would suffer enormously.

South Carolina

FOX News (Fivethirtyeight rating B)

Clinton 65

Sanders 21

O’Malley 3

FOX hasn’t polled in South Carolina yet, so we don’t have any apples to apples comparisons to make, but this is within the range of other polling we’ve seen in the state. Clinton’s 44 point lead is slightly weaker than the 49 point lead she holds in the RCP and Pollster averages, but not significantly.

The poll shows Sanders’s continuing difficulty connecting with Democratic base voters.  Clinton leads with black voters 82-11, with women 72-17, voters making less than 50k 71-17.

Sanders’s best demographic continues to be white voters, where Clinton only leads 39-37. He’s “only” back by 27 points among men, and only back by 33 points among liberals, his other best demographics. In other words, Sanders is losing badly across the board except with white voters.

New Jersey

Rutgers/Eagleton Poll (Fivethirtyeight rating C+)

Clinton 60

Sanders 19

O’Malley 1

New Jersey comes late in the process, on June 7th, so it is unlikely that there will be much impact from this race on the overall nomination. That said, the data is out there, so here it is for your perusal. Their last poll two months ago was 49-19-10 (Biden included) so basically Clinton has gained Biden’s supporters and Sanders has not moved.

For what it is worth, there have been a couple of other polls in New Jersey since the beginning of November, and they all show pretty much the same race. In them Clinton led 64-27 and 56-23. Given Clinton’s time as Senator from New York, and her prominence in general, and the fact that the state likely won’t get much attention I don’t think we can read too much into the numbers here.

NATIONAL

CBS/New York Times (Fivethirtyeight rating B)

Clinton 52

Sanders 32

O’Malley 2

This poll shows almost no change from their poll a month ago, which showed a 55-33 race. Obviously having a stable twenty point lead for the month going into the holiday season is good news for Clinton.

Clinton’s support is also more committed than Sanders’s, with 58% of Clinton supporters saying they have made up their minds, while just under half of Sanders’s supporters are committed at 47%.

The good news for Sanders, if you want to find some, is that, though this poll shows stability, the bulk of the polling IS starting to center closer to a lead in the low 20s for Clinton rather than the 25-30 point leads we were seeing a couple of weeks back.

YouGov/Economist (Fivethirtyeight rating C+)

Clinton 56

Sanders 35

O’Malley 2

Not a huge difference from their poll two weeks ago, which had the race at 54-34-4. What we’ve been seeing pretty consistently across polls is a fairly steady race, with Clinton in the mid to high 50s and Sanders in the low to mid 30s.

(Pollster is currently reporting this as the 63-30-2 race that we see among Democrats- excluding Independents- but I assume they will update it to the metric I used above, which is Democratic Primary Voters, which is reported for their previous poll. For context, the previous poll among all Democrats was 64-26-3, so a slight improvement for Sanders among Democrats)

Public Religion Research Institute (Unrated by Fivethirtyeight)

Clinton 52

Sanders 31

O’Malley 4

This poll resembles other recent polling, but it does represent a significant shift towards Sanders over their last poll a month ago that showed Clinton up 62-29. For some reason, however, their November survey did not include Independents who lean to the Democratic party. For that reason, there isn’t much value in comparing the numbers, besides to note that they confirm what we’ve seen elsewhere- Clinton dominates among Democrats while Sanders does better with Independents. As many states allow people to choose their primary (or register with the party of their choice in order to participate) we can assume that this month’s results are closer to what we’re likely to see when actual voting starts.

The hidden danger for Sanders, though, is that if Independents start to see the Republican race as more competitive or interesting he may be losing one of his strongest voting blocs to the other primary. As we discussed in the New Hampshire poll above, that difference could be key.

Gravis Marketing/One America News (Fivethirtyeight rating C)

Clinton 62

Sanders 30

O’Malley 9

If anyone cares what Gravis has to say about the race, this is an improvement for Clinton over their last poll, where she lead 59-32 near the end of November. I don’t know that the C rating does the crappiness of Gravis justice, but we’ll throw it on the pile with the rest.

Ipsos/Reuters (Fivethirtyeight rating C+)

Clinton 56

Sanders 30

O’Malley 5

The “official” weekly release of the daily tracker comes today, and with it the wild and unpredictable swings we’ve come to know and love from this poll. These results are a composite of their last five days of interviews, and they release numbers each Wednesday (though daily results are often available on their site)

Their last Wednesday release was extensively diaried on DailyKos as it showed a shockingly close 51-36 point race. Those of us who have been following the poll cautioned against reading too much into that 15 point race, and you can see why. The race has (allegedly) shifted by 11 points in Clinton’s favor in the last week. This is no more plausible than the huge shift in Sanders’ favor before that. Take a look at their most recent releases:

Nothing in the real world has been happening that would provide a compelling explanation for these shifts. There doesn’t seem to be any consistent lean in the numbers, either. Sometimes they are better or worse than contemporaneous polling for either candidate, sometimes both, sometimes neither. Taken with a HUGE grain of salt, we can throw this poll on the pile with the rest.

THE AVERAGES

All numbers are Clinton-Sanders-O’Malley

New Hampshire

RCP: 43-47-4 (Sanders +4)

Pollster: 44-45-3 (Sanders +1)

South Carolina

RCP: 70-21-3 (Clinton +49)

Pollster: 71-22-2 (Clinton +49)

New Jersey

RCP: 60-23-2

Pollster: 63-23-2

National

RCP: 55-31-2 (Clinton +24)

Pollster: 55-31-3 (Clinton +24)

Overall:

A lot of new data that doesn’t change the overall picture much. We stay tied up in New Hampshire, South Carolina looking out of reach for Sanders, and Clinton with a strong and steady lead nationally. With just two weeks to go until Christmas, there is a limited window in which candidates can make a major difference until next year.

THE HISTORY

THEN numbers are taken from the five most recent polls starting on November 11th 2007 (53 days before the Iowa Caucuses). NOW numbers are taken from the five most recent polls starting today (53 days before the Iowa Caucuses)

New Hampshire

Then

Clinton 32

Obama 23

Edwards 14

(Clinton +9)

Now

Clinton 39

Sanders 47

(Sanders +8)

As we’ve seen before, New Hampshire remains Sanders’s strongest bet. To give you a sense of how different the poll released today is from the rest, excluding that poll from the others and giving us an average of the last four polls gives us a 45-46 race. Either way, New Hampshire is the place where Clinton is outperforming her 2007 campaign by the least, and where Sanders is outperforming Obama by the most. In the 2008 race, Clinton held steady leads in New Hampshire until the Iowa caucuses, trailed in almost all the polling after Iowa, but went on to win the state. At the time, Clinton outperformed her polling by 9 points while Obama underperformed his by about 2. New Hampshire is a key state- if Sanders wins there he may be able to bring some momentum out of it, and Clinton might be able to end his campaign by winning there.

South Carolina

Then

Clinton 40

Obama 25

Edwards 12

(Clinton +15)

Now

Clinton 68

Sanders 21

(Clinton +47)

While then-Senator Obama WAS trailing at this point in 2007 in South Carolina, it was by a significantly smaller margin. In fact, Obama’s share of the vote at this point was higher than Sanders currently has.

Sanders’s continuing weakness here is also what was Obama’s hidden strength- African American voters. After winning Iowa, President Obama took a huge lead among African American voters, and won them on election day 78-19, partially on the strength of the historic nature of his candidacy combined with increased confidence in his ability to win after winning over the overwhelmingly white voters in Iowa. For contrast, the 59 point lead Obama had on election day among black voters that powered him to a huge 29 point win in the state is significantly smaller than the 71 point lead Clinton has over Sanders today among this group. Sanders also has remarkably high unfavorable ratings with South Carolina Democrats (particularly African Americans), meaning it is unlikely that he will be able to ride a huge wave of momentum into South Carolina even if he wins Iowa or New Hampshire (or both).

New Jersey

Then

Clinton 48

Obama 20

(Clinton +28)

Now

Clinton 57

Sanders 22

(Clinton +35)

For the now numbers, we only included four polls, rather than five, because the fifth poll would have been from back in mid April, when Sanders scored a whopping 0%, as did every other candidate besides Clinton (since no one was running). In 2008, New Jersey was a Super Tuesday state, as opposed to its slot at the end of the process this year. Not much use comparing them besides to say that Clinton’s lead average was down to single digits after Iowa, so there is room for improvement for a candidate who can win early even in a state where the frontrunner is very well known.

National

Then

Clinton 47

Obama 23

(Clinton +24)

Now

Clinton 56

Sanders 30

(Clinton +26)

Clinton had a particularly strong stretch of polling at this point in November of 2007, leading us to a similar margin between the eight year stretches. The primary (ha) difference remains that Clinton is solidly over 50 and there is a much smaller pool of voters undecided or supporting another candidate. While Clinton’s lead is of a similar margin, it is of a different quality.