All Tied up 47 to 47 in New Hampshire — Granite State Poll

New Hampshire is Romney’s last stop on the campaign trail and it may just be the state to put him over 270.  The latest from the WMUR Granite State poll shows a dead heat at 47 with 4% Undecided.  This is the umpteeth poll where the President fails to get above 50%, an unquestionably bad sign for any incumbent.  Romney leads among Independents by 22-points 54 to 32.  Winning the Independents is how you win New Hampshire so quite honestly I’m surprised it’s tied … but by now we know what that means.  Unfortunately no Party ID was given. Thanks to MikeP’s eagle eyes, we see the party ID is D +3 (Dem 42, Rep 39, Ind 19). In 2008 it was also D +2 (Dem 29, Rep 27, Ind 45) and in 2004 it was R+7 (Dem 25, Rep 32, Ind 44).  Note the consistently high % of self-identified Independents in 2008 and 2004. Now we know the Granite State Poll has far too few Independents and even too many Democrats. If Romney solidly wins Independents this state is turning red on November 6:

Mitt Romney has closed the gap with Barack Obama in the Granite State, and the two are now locked in a dead heat. In the most recent Granite State Poll, 47% of likely New Hampshire voters say they will vote for Obama, 47% say they will vote for Romney, 2% prefer some other candidate, and 4% are undecided. When undecided voters are asked which candidate they lean toward, Obama and Romney remain tied, with 48% for Obama, 48% for Romney, 2% for some other candidate, and 3% remain undecided. Obama has lost considerable support since 2008 as only 87% of 2008 Obama voters say they will vote for him this time while 94% of McCain voters say they will vote for Romney. Romney runs strongest in the vote rich Manchester area and in towns along the Massachusetts border while Obama gets more support in the North Country and in the Connecticut River valley. There is a pronounced gender gap, but neither candidate benefits. Obama leads among women by 58% to 40% while Romney leads among men, 57% to 38%.

Both candidates have secured their bases — Obama currently has the support of 95% of New Hampshire Democrats, and Romney has the support of 95% of Republicans. Independents are breaking heavily to Romney — 54% say they will vote for Romney while only 32% say they will vote for Obama. “The movement of Independents to Romney is the most significant factor in his making this race a dead heat,” said Andrew Smith, Director of the UNH Survey Center. “Two weeks ago, independents were divided in who they would support.”

For President Percent
Barack Obama 47
Mitt Romney 47
Other 2
Undecided 4

78 Comments

  1. Brad
    Posted November 3, 2012 at 9:45 pm | Permalink | Reply

    And lookie here…more good news when you realize turnout is going to be about D+4-5:

    NumbersMuncher ‏@NumbersMuncher

    PPP(D) finds Obama up 6 in PA, 52-46. Sample is D+10 — it was D+7 in 08 and D+2 in 2010. Obama needs close to 08 turnout to win even in PPP

    • Interested Party
      Posted November 3, 2012 at 9:52 pm | Permalink | Reply

      Anyone who quotes the ppp poll MUST also use the Susquehanna poll as well. Coming out tomorrow.

    • Brad
      Posted November 3, 2012 at 9:55 pm | Permalink | Reply

      I’m quoting it because if you take out +6 for Dems and re-weight you get a much tighter margin.

      • Interested Party
        Posted November 3, 2012 at 10:14 pm | Permalink

        I meant some Poll troll as evidence that PA is not in play, not you, of course!

      • Brad
        Posted November 3, 2012 at 10:22 pm | Permalink

        Oh…a preemptive strike…I like it. Carry on.

  2. Kyle
    Posted November 3, 2012 at 9:46 pm | Permalink | Reply

    Redrew the map with a Romney win in MN and NH. If he were to take these two states plus CO, he could lose OH and WI and still win with 271.

    • nvClark
      Posted November 3, 2012 at 9:51 pm | Permalink | Reply

      Yes, but if he were to take MN and NH, why would he lose OH and WI? Are you ready for the landslide?

      • Brad
        Posted November 3, 2012 at 9:56 pm | Permalink

        Yeah, he would at least win WI given his money advantage there.

      • nvClark
        Posted November 3, 2012 at 10:09 pm | Permalink

        Brad, how come not OH since it’s been polled as a dead heat for so long, and is generally considered much more likely than MN and WI? So much so that when Romney went into those states it was portrayed by Dems as a desperation move. OH polling hasn’t moved much but I think that’s because it is being held. OH is a black box at this point but if MN and WI go then how could OH not? to me that is really hard to imagine.

      • Brad
        Posted November 3, 2012 at 10:25 pm | Permalink

        I have this semi-irrational notion that O is about even in OH. Call it the ghost of 04 past haunting me where Kerry significantly outperformed the national average in that state. OH has thrown the kitchen sink at it this cycle and I just think that even with R+5 nationally, it could still be amazingly tight due to OFAs ground game in OH. I want at least a national poll, even a Rass poll showing +2-3 before I bite.

      • Brad
        Posted November 3, 2012 at 10:30 pm | Permalink

        …meant to say too that O is less organized and more vulnerable in MN, WI and PA due his over-focus on OH and Romney killing him right now in ad $s in WI, MN, PA and I think MI. Plus, I think R has a big ace in his back pocket with WI, as conservative voters there have repeatedly gone to the polls each fall over the past 2 years due to recall efforts. The WI infrastructure, as we saw in the Walker/senate reps recall last fall and summer of 2011, is equal if not better than Dem/OFA this year. Of course, all this is just my gut talking.

      • nvClark
        Posted November 3, 2012 at 10:42 pm | Permalink

        I understand where you’re coming from but considering the focus on OH, conceding that Romney was ahead there would be tantamount to conceding the election. That they simply will never do. So personally I think OH polling is completely useless rubbish. Particularly considering the claims I’ve read in other threads (which I did not verify) That Obama won Ohio by over 200,000 votes in 2008 and the Reps lead Dems in early voting in Ohio this year by over 200,000…Which would completely erase that lead. (without even considering the swing in turnout and enthusiasm that such an early voting lead would imply for election day turnout)

      • Brad
        Posted November 3, 2012 at 10:56 pm | Permalink

        I like you’re thinking nvClark….I hope we are both laughing at my expense on Tues.

      • stuckinmass
        Posted November 3, 2012 at 11:08 pm | Permalink

        there’s a lot of concern over OH, the “obama poisoned it” theory

        I think people are just looking for evidence that Rr could take PA/WI/MN or even MI to avoid needing OH

  3. damien
    Posted November 3, 2012 at 9:46 pm | Permalink | Reply

    wasnt obama up 15 in this thing?

    • Posted November 3, 2012 at 10:06 pm | Permalink | Reply

      Obama was up 15 but that requires a Dem +25 sampling ROFLMAO GO R&R GO!

  4. Posted November 3, 2012 at 9:48 pm | Permalink | Reply

    Obama won indies by 20 in 2008. If Romney leads indies by 22 it’s over in NH.

    • damien
      Posted November 3, 2012 at 10:13 pm | Permalink | Reply

      thats what i dont get…95 a piece but romney up 22 among indies but its tied?

      • No Tribe
        Posted November 3, 2012 at 11:29 pm | Permalink

        Landslide. This is how NH rolls, always breaks late and breaks hard.

    • Brad
      Posted November 3, 2012 at 10:31 pm | Permalink | Reply

      Yup. And O won the state by 10.

  5. William Jefferson Jr.
    Posted November 3, 2012 at 9:52 pm | Permalink | Reply

    If both get 95% of their base, respectively, and Romney leads by 20 points among indies, how do they end up tied? The independent number must have been pushed pretty low, coupled with a Dem oversample.

    • tim
      Posted November 3, 2012 at 9:55 pm | Permalink | Reply

      two words – dem oversample.

      Look, all, this wont be a landslide. But it could very well be a decisive Romney victory, and a slap in the face to the MM and its minions.

      Vote and get involved. Two days left.

  6. MikeP
    Posted November 3, 2012 at 9:53 pm | Permalink | Reply

    The party ID is roughly D+2 in this New Hampshire poll(UNH-WMUR). It’s funny the last UNH poll showed Obama up by 9 (the party was way off). David Messina quoted the UNH as “evidence” that Obama was going winning New Hampshire.

    The college pollesters giveth and the college pollestes taketh away….

    • William Jefferson Jr.
      Posted November 3, 2012 at 9:56 pm | Permalink | Reply

      D+2? If Indies were at their historic average–45% of the electorate–Romney should have wiped out the partisan ID deficit, no? Did they push indies down into the teens?

      • MikeP
        Posted November 3, 2012 at 9:59 pm | Permalink

        oops I mean Jim Messina,

        According to my math ( I”m getting a democrat party Id of roughly 41% and a repulican party id of roughly 39%), leaving 20% party ID as independent.

      • MikeP
        Posted November 3, 2012 at 10:01 pm | Permalink

        http://www.unh.edu/survey-center/news/pdf/e2012_pres110312.pdf
        party Id is on page 7 of 22

      • Posted November 3, 2012 at 10:05 pm | Permalink

        I knew to look there and still missed it. I’m just tired. Thanks for catching that.

    • Brad
      Posted November 3, 2012 at 9:57 pm | Permalink | Reply

      Didn’t their last poll show D+6…funny how it tightens when you reflect a more reasonable turnout assumption.

      • MikeP
        Posted November 3, 2012 at 10:03 pm | Permalink

        D+2 doesn’t seem that reasonable to me…. I see more like R+3 as reasonable, however I doubt Romney is winning independants as strongly as this poll suggest ( nevertheless I think he is winning independants ).

  7. stuckinmass
    Posted November 3, 2012 at 9:55 pm | Permalink | Reply

    with the high number of I and the historically close D+R, how is Romney not ahead?

    • Brad
      Posted November 3, 2012 at 9:58 pm | Permalink | Reply

      Shhh. He is. 🙂

  8. Ranger375
    Posted November 3, 2012 at 10:01 pm | Permalink | Reply

    It is amazing to me how different an election looks when looking at things as you all do. Through your eyes there is so much more clarity.

    Thanks to all of you!

  9. Brandon
    Posted November 3, 2012 at 10:03 pm | Permalink | Reply

    I think independents were about 20% in this poll, far too low in my opinion for a traditionally independent state.

    • Brad
      Posted November 3, 2012 at 10:05 pm | Permalink | Reply

      Well there’s your problem…2008 exit polls had D/R/I at 29/27/45

      • Brad
        Posted November 3, 2012 at 10:05 pm | Permalink

        Oh…and linky:

        http://elections.nytimes.com/2008/results/states/exitpolls/new-hampshire.html

      • Brandon
        Posted November 3, 2012 at 10:08 pm | Permalink

        Correct, which would seem a tie is conservative bases on this polls internals

      • Brad
        Posted November 3, 2012 at 10:12 pm | Permalink

        Actually, you’re too kind. I was thinking it was more of the laughing hysterically at WMUR type of thing…. Just calculating the Indies (45% Indies at +22R) alone would give Mitt a 10% cushion before you reconcile the rest of it…if D/R is split then it’s +10 – or you would need a boatload of R crossovers or most of the Rs to stay home for O to win with an Indie break like this…. Still, with all flaws accepted, it’s a great poll for Mitt.

      • NHConservative
        Posted November 3, 2012 at 11:04 pm | Permalink

        I can tell you this, R’s aren’t staying home here on Tuesday. I already voted abesentee in the event I get hit by a bus before Election Day.

  10. Neil in NC
    Posted November 3, 2012 at 10:05 pm | Permalink | Reply

    I would love to see Indies break for GMR in Nevada by 1/2 this amount.

  11. kenberthiaume
    Posted November 3, 2012 at 10:06 pm | Permalink | Reply

    there’s two “party ID” questions, one is by registration (which I think is like 30/30/40) and one by “which do you identify with” which was D+2 in 08, I believe.

    They asked people who voted before “who did you vote for” and it was Obama 55, mccain 45 approx, which was very close to the actual vote in NH. So it is interesting. See who actually shows up I guess.

  12. Ken in Bama
    Posted November 3, 2012 at 10:07 pm | Permalink | Reply

    Interesting thing about Nate Silver. He uses multiple polls from the same organization…so wouldn’t that multiply a problem if the particular pollster has a PVI that is biased? Plus he has a marist poll from 5/20 that he still gives (though slight) weight to. And everything from 8/21 on seems to be counted and weighted. In this scenario, if 20 polls were released this weekend saying Romney was up by 1-2 in Ohio…I am not sure even that would be enough for Nate’s “model” to project a Romney victory. I am being a little overly-dramatic here, but I think you get my point.

  13. dingMan
    Posted November 3, 2012 at 10:07 pm | Permalink | Reply

    Don’t forget to change your clocks tonight and your President on Tuesday.

    • Ken in Bama
      Posted November 3, 2012 at 10:11 pm | Permalink | Reply

      Hehee…my wife said that to me tonight.

    • Brad
      Posted November 3, 2012 at 10:14 pm | Permalink | Reply

      Yeah, I’d like to set my clock back one administration.

  14. Derclaw86
    Posted November 3, 2012 at 10:08 pm | Permalink | Reply

    The poll says both sides have secured their respective base voters. If Romney is up by over 20 points among Independents, how can this poll be anywhere close to being tied? If Independents are over 40 percent of the NH electorate, and Romney is winning this group by 54-32, it would take an almost D+10 sample just to get this to a tie. Something is really screwy with this poll unless we believe that Obama’s NH prospects are free falling. (Like Granite maybe!)

  15. TheTorch
    Posted November 3, 2012 at 10:12 pm | Permalink | Reply

    Oh dearie, dearie me. This might be the saddest rally yet for Obama!

    Cleveland Ohio rally for Obama – Get Out The Vote – featuring Stevie Wonder!

    attracts 200 people!????

    Someone ring the whitehouse and tell them the party is over!

    Apparently they are blaming lack of information i.e. people not knowing in advance.
    But according to one of the commenters they knew about it 2 days in advance! LOL.

    • TheTorch
      Posted November 3, 2012 at 10:13 pm | Permalink | Reply

      oops, should of included link:
      http://www.cleveland.com/metro/index.ssf/2012/11/few_know_about_stevie_wonder_a.html#incart_river

      • JGS
        Posted November 3, 2012 at 10:49 pm | Permalink

        So much for their vaunted organizational skills.

    • Brad
      Posted November 3, 2012 at 10:16 pm | Permalink | Reply

      At first, I thought you were lying. Then I saw your link. Then I read this:

      “Fewer than 200 people showed up to watch Wonder perform a handful of his hits at the early voting event in support of President Barack Obama.

      Most of the people had learned of the event just hours before the 9:30 a.m. start time. Some just happened to be in the area and followed the live music.”

      Wow.

      • TheTorch
        Posted November 3, 2012 at 10:19 pm | Permalink

        I didnt believe it when I first saw the link somewhere else! But it is true! LOL

      • Bob San Diego
        Posted November 4, 2012 at 2:05 am | Permalink

        I would have showed up to watch Stevie Wonder.

        What a voice.

  16. Posted November 3, 2012 at 10:12 pm | Permalink | Reply

    my back of envelope calc’s would indicate their model as:

    Dem 37.6%
    Rep 30.6%
    Ind 31.8%

    I derive that on excel by assigning 100k each to Dem & Rep and then using the 54/32 Ind split backing into vote total necessary to give a 48/48 tie

    I am sure I am not quite ‘exact’ but seems like Dem + 7.0% which corresponds to what others have above

    • Brad
      Posted November 3, 2012 at 10:18 pm | Permalink | Reply

      Yeah, and the Indy split is based on 96 total respondents. The MoE is just way to high to be useful for this….

      • Brad
        Posted November 3, 2012 at 10:20 pm | Permalink

        Okay, I had to check. MoE is +/- 10 on an N=96 assuming 700,000 voters (2008 assumption). Can I laugh at WMUR now?

      • O'BuhBye
        Posted November 3, 2012 at 10:56 pm | Permalink

        Should be 1/(n^.5) no matter population size.

  17. stuckinmass
    Posted November 3, 2012 at 10:15 pm | Permalink | Reply

    i rebalanced the numbers with something more reasonable 43 I, 30 D, 27 R – kept the D+ 3 Romney would be up 1

    • Brad
      Posted November 3, 2012 at 10:40 pm | Permalink | Reply

      That doesn’t look quite right. I get 50-44 when using 2008 assumptions.

      • Brad
        Posted November 3, 2012 at 10:41 pm | Permalink

        50-44R that is..

      • stuckinmass
        Posted November 3, 2012 at 10:53 pm | Permalink

        my spreadsheet could be wrong.

  18. Posted November 3, 2012 at 10:17 pm | Permalink | Reply

    Obama will lose Ohio. Yes that is right. Obama will lose Ohio

    Here is why:
    According to Obama supporter in Ohio: “There is significantly less enthusiasm.” “…Less young people volunteering like 2008″

    http://www.politico.com/news/stories/1112/83267.html?hp=t1_3

    • TheTorch
      Posted November 3, 2012 at 10:34 pm | Permalink | Reply

      Yes interesting article and this paragraph struck me:

      Of the Romney campaign’s suggestion that the early vote strategy is simply decreasing their election day share, Ohio Obama director Pickrell replied: “I fail to understand why you wouldn’t bank as many votes as you can.”

      I thought that was a WOW moment, it confirms they are cannibalising their early voting. The Obama guy, did not even challenge the question.

      Confirms what Karl Rove has stated.

  19. Tony
    Posted November 3, 2012 at 10:18 pm | Permalink | Reply

    Des Moines Register has their final Iowa poll with Obama at 47% and Romney at 42%. Their final 2008 poll had Obama +17, he won by 9. Just saying….

    • TheTorch
      Posted November 3, 2012 at 10:22 pm | Permalink | Reply

      Yes apparently this pollster for Des Moines / Iowa, is inaccurate, has a poor record, they had the 2004 wrong as well!

      • Tony
        Posted November 3, 2012 at 10:26 pm | Permalink

        Yea, you are exactly right.

    • Posted November 3, 2012 at 10:27 pm | Permalink | Reply

      I just looked at RCP recent polls. Avg. + O 2.5% moe ~ 3.5/4.5 or so

      two thoughts:
      1) Obama still not above 50% for state he won by 9 – not good for him
      2) I seem to remember the register stating their endorsement meeting was rather feisty – who’d bet that one of losing Editors (wanting to endorse O) directs the ‘slant’ of the Dem/Rep/Ind such that it is D+6 or better

      • trux
        Posted November 3, 2012 at 10:34 pm | Permalink

        The editors have nothing to do with the internals of the poll.

    • Posted November 3, 2012 at 10:38 pm | Permalink | Reply

      Des Moines Register poll is D+5. 2008 turnout was D+1, 2010 R+1, current registration is R+1 I believe

  20. Shane kovac
    Posted November 3, 2012 at 10:23 pm | Permalink | Reply

    Too Many Ds and too many Rs not enough Is

    Still bad for O good for R/R

    Keith and you all have done more to pep me up today I swear. Was explaining this site and the numbers to my dad at dinner.

    He is very confident when I said my gut was leaning Obama to win by a sliver thought he was going to slap me out of the car. He owns a gun store, and jewelry store and pawn shop. Has a ton if cop customers he said unanimously they were alike Screw our union we are going Rimnety

  21. Jan
    Posted November 3, 2012 at 10:34 pm | Permalink | Reply

    IOWA EV update:

    Ballot requests:
    Dems 301,767 (42.60%)
    Reps 219,354 (30.97%)
    Inds 186,262 (26.30%)
    TOTAL 708,313

    Ballot requests today (first time Reps requested more ballots):
    Dems +5,690
    Reps +6,132
    Inds +6,087

    Ballots cast:
    Dems 261,166 (42.55%)
    Reps 198,130 (32.28%)
    Inds 153,684 (25.04%)
    TOTAL 613,749

    Ballots cast today:
    Dems +10,088
    Reps +9,832
    Inds +9,692

    So with only a 256 ballot gain today, dems are now leading with 63,036 ballots. With only the tiny leads they are getting every day, they are not close to reaching the 70,000 gap that I argued should be quite easy to overcome on election day assuming a 50/50 Indie split.

    • TheTorch
      Posted November 3, 2012 at 10:40 pm | Permalink | Reply

      Very very encouraging.

      • Brad
        Posted November 3, 2012 at 10:44 pm | Permalink

        Yeah I think RR said it would take a gap of 130k to overcome what they’re anticipating on election day.

      • Jan
        Posted November 3, 2012 at 11:04 pm | Permalink

        I’d say that number isn’t unreasonable. For the sake of argument and I assumed turnout and crossover being equal to 2008 and albeit a 15 point swing from 2008, just a 50/50 split among indies. I’d say all or most of those variables will probably turn out better..

    • kenberthiaume
      Posted November 3, 2012 at 10:59 pm | Permalink | Reply

      It seems like there are still about 95K absentee ballots outstanding. Could mean a delay in the vote totals if those result in provisional ballots.

    • No Tribe
      Posted November 3, 2012 at 11:32 pm | Permalink | Reply

      Thanks Jan.

  22. Tony
    Posted November 3, 2012 at 10:40 pm | Permalink | Reply

    There are going to be shocked people on November 7.

    • JGS
      Posted November 3, 2012 at 10:53 pm | Permalink | Reply

      Including Jon Ralston, whom I used to respect but now know to be a total partisan tool hack.

  23. Guest
    Posted November 4, 2012 at 4:51 am | Permalink | Reply

    Why are NH indies so fickle?

  24. UncleFred
    Posted November 5, 2012 at 1:05 pm | Permalink | Reply

    The NH poll is complete junk. Lets look at some cross tabs.
    35% of NH voters have a college degree or better. 60% of this sample have a college degree or better of whom just over 1/3rd have an advanced degree. The gender gap reflected in this poll is a result of over sampling college educated women in the state who tend to be liberal and whose votes are heavily influenced by the contraception issue. Lastly the largest partisan demographic is Republicans who have about 2% more than Democrats who are narrowly trailed by independents. Say 35%-R 33%-D and 32%-I. Assuming that the lead among independents is accurate, Romney would be up outside the MOE.

    That said both campaigns believe that the race in NH is +-2 maybe 3, and that it will all come down to turnout.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: