« BACK  |  PRINT

RS

FRONT PAGE CONTRIBUTOR

Swingometer: Gallup Party ID figures predict solid Romney win

We always talk about the independent, swing vote in elections because those tend to be the persuadables. But party ID numbers matter as well, because those partisan voters tend to split better than 90/10 for their party.

It is for that reason that Gallup’s new partisan ID split, one that mimics what Rasmussen has been saying all along, predicts nothing less than doom for the Democrats, and a solid, national win for Mitt Romney this year.

Ignore Gallup’s headline. They’ve buried the lede so far deep, they’ll be fracking in Australia to bring it to the surface. Demographically the country hasn’t changed in a few years, naturally, but the difference between 2008 and 2012 is in the TEA party. The TEA party happened, dictated the 2010 elections, and has now resulted in a large partisan ID and registration shift in this country since.

When Rasmussen Reports noted this, and showed it in polling, Nate Silver and his followers scoffed. Just as they did in [2010] when Rasmussen was first to predict the TEA party driven Republican wave, they criticized and insulted the firm that was first to what turned out to be the facts.

Now Gallup is in the game, and the numbers are brutal. In 2008, the Democrats had a 39-29 (D+10) advantage in hard party ID, and a 54-42 (D+12) advantage with leaners. In 2012 though, we’re in the post-TEA party era. Republicans now show a 36-35 (R+1) hard party ID advantage, and a 49-46 (R+3) lead with leaners. This gives us a range of party ID swings from 2008, from R+11 to R+15.

That Gallup is giving us a range is not new. They did the same in 2010, when the late Generic Ballot polls offered two different voter models. First there was the high turnout model, one that proved too pessimistic for Republicans, and the low turnout model that was too optimistic. The truth was, as they surely intended, somewhere in the middle.

But let’s look at where this range takes us. Asking Swingometer what an R+11 swing means, it’s bad for Democrats. This is a crude measure as Swingometer is meant to deal with the two party vote and not raw party ID, but again, with the 90/10 split that partisans show in their candidate support, this isn’t going to be that far off.

This low end swing gives Romney the election handily. We get taken back to the 2000 election map, but with Romney trading Nevada in exchange for Iowa, Minnesota, and Pennsylvania. It’s a solid 315-223 Electoral College victory. Not a blowout, but roughly as meaningful as Obama’s 2008 victory.

But then we come to the high end, R+15 swing:

Again, not a blowout. The nation is more polarized than it was in the past, so elections like 1984 or 1964 just don’t happen these days. but gaining the straight 2000 map plus Minnesota, Wisconsin, Iowa, Pennsylvania, and Maine’s second district would give Romney a 332-206 victory.

The takeaway here is that Mitt Romney has many paths to victory. He’s solidified enough states (Florida, North Carolina, Virginia, Indiana, and probably Colorado) that he has his chances elsewhere (Pennsylvania, Ohio, Wisconsin + Iowa or New Hampshire, Nevada + New Hampshire + Maine 2).

This election is winnable for Mitt Romney as long as his people vote and get out the vote on election day.

COMMENTS

  • earlgrey

    Thanks for the headline Neil. How dd you know I QA having a bad day?

    I am not even going to bother reading all the disclaimers in the text. I am just going to bask in the headline. Hey, it beats drinking.

  • deano64

    Nice Neil. I’ve been missing your poll analysis this election year. Hope to see more.

  • tyman

    Neil, I haven’t had a chance to read your post yet, but I have missed your info. Do you know if Rasmussen has changed how he samples D/R/I? The last I heard, he was using a D +3 sample.

  • maddog

    Given public polling, one would think that Wisconsin would shift before Minnesota and Michigan before Pennsylvania. But enough quibbling. If this shift ensues as predicted, the Great Blue Wall of 1992-2012 will crumble, and bring with it enough states to handily put Mitt Romney in the White House. Make it so.

  • http://www.hakubi.us/ Neil Stevens

    I’d have to look them up.

  • http://www.hakubi.us/ Neil Stevens

    Thanks. I took a full time job in April (to get out of California), and that killed my polling posting for now. Just don’t have the time with a 2.5 hour commute, one way ( so 5 hrs total), 2-3 days a week.

  • http://www.hakubi.us/ Neil Stevens

    I suggest you read JUST the last line then :)

  • canadianlawyer

    The real key is: Does anyone actually believe that the election will look like this?

    Both of those models have Romney winning Iowa. Romney has already lost Iowa. Any democrat or republican believing that either party is going to hit 310+ electoral votes is really drinking the strong kool-aid.

    Ohio, Wisconsin, New Hampshire, Nevada and as any election, likely one surprise state. Anything else is far too optimistic for its own good.

  • gawken

    Neil..I am now thoroughly confused. This past weekend, the head of Gallup was on FOX NEWS Sunday with Chris Wallace. When asked about polls, specifically all the attention paid to the “sampling” done in recent polls, the D/R/I ration, whether this election mirrors 2008, or 2010, he replied that Gallup does NOT, nor has it ever, asked those whom it polls about their party affiliation. They consider it irrelevant. Wallace seemed as astonished as I was at that response, but did not ask any follow-ups. The Gallup guy did say that at the end of the poll, they do ask the question, but it doesn’t factor into their results. So why do they ask it? And here we are saying, nay, rejoicing, that Gallup’s data on this is meaningful, and it’s a good thing. ( sorry for the Martha Stewart impersonation.)

    I’m ecstatic that the only poll that actually counts comes in a few more days, but if you could provide some guidance, some insight, it’d be appreciated.

  • Veronica

    Yes, Neil, I’d been waiting for you to show up w/ your polls for awhile, too. Drudge’s updates just weren’t cutting it anymore. Glad you’re out of Cali, I wish you good fortune with your job.

  • http://www.hakubi.us/ Neil Stevens

    Hearing the commentary second hand I can only speculate. but it sounds like what Gallup was really saying ther is that they don’t post weight by Party ID.

  • Darin_H

    I’d figure at least a 14 point swing right now – Obama won WI by 14 and they’re tied right now.

  • gbenton

    To the point that Neil made in the post, the Tea Party happened. If you look at the way Democrats got overrun in 2010 – the last national election – the real world results were too optimistic, too… but that’s what happened.

    Think of this like a market correction after a bubble or mania. The crash is epic, not normal 10 to 20%. Obama has been a mania in that it was irrational exuberance that got him elected, not ‘fundamentals’.

    So when his bubble got pierced on October 3rd, the preference cascade began for Romney and I believe a frustrated nation will vent their anger towards him in a way that will astonish you.

  • RyJohn

    Neil/gawken: Hugh Hewitt recently interviewed some pollsters and pressed them on that same question. The short answer is that pollsters consider it irrelevant. They weight age, gender, race, education and income (among a few others). Those characteristics are more static and easier to measure whereas party affiliation is in a constant state of flux (and thus a less meaningful metric).

    Changes in party affiliation, they claim, reflect mood changes in the electorate. So, if their sample shows a D+9, for example, that is reflective of voter attitudes that favor democrats. Although unlikely, a poll with party identification of, say, D+30 is considered reliable, assuming the other static metrics were properly accounted for.

    It makes little sense to me. But then again, I take all polls with a grain of salt because the response rate is so low. When a pollster has to make 30,000 calls to get a sample of 1,000 RVs or LVs, the sample is questionable.

    Check out IowaHawk’s spot-on (and hilarious) exposition on this topic. http://iowahawk.typepad.com/iowahawk/2008/10/balls-and-urns.html

  • http://rightwingbuckeye.wordpress.com/ Right-Wing Buckeye

    Andrea Tantaros just mentioned this post, and you by name, on The Five. Congratulations.

  • runner12

    Hey Neil! They just gave you a nod on The Five on Fox. Great analysis.

  • lineholder

    Neil, FWIW, you also got linked in at RCP (afternoon edition).

    Just curious, but have you checked this out?
    http://elections.gmu.edu/early_vote_2012.html

  • swanman823

    This makes sense. And beyond what seem like irrational polls I think you see and feel in both liberal and conservative comments a wave building towards Romney. The fear and bitterness in the Presidents speaches and eyes is palpable. Chris Matthews is going to be locked up in the same padded cell they put Keith (remember him) Olberman into. The libs are melting down, it’s all racism now. I believe there will be a comfortable Romney win and watching MSNBC and the MSM in general is going to be a lot of fun on election night.

  • paulinpittsburgh

    When all is said and done and Romney wins 320 EC votes can we dare to hope that Nate Silvers pathetic misapplications of statistical modeling will finally tossed in the trash bin where it belongs? The only other people Silver I’ve seen paid to do a job they’re more frequently wrong at than right are local weather forecasters.

  • Dave_A

    Well, my vote for Romney has been in since September (WA allows *really* early voting for anyone with military ID)…. Not lot of good it will do, living in this single-party-rule state…

    Just have to hope that Soros’ gal doesn’t get elected sec-state, and that maybe we can get a squish-Repub as governor (a monumental feat in WA)….

  • Dave_A

    Also, this is way, way optomistic…

    269-290 is a far more believable vote-range, all things considered…

    If Romney does NOT get Ohio, 269 is our only hope (and that involves winning CO, NV, IA, and NH – since WI, MI, and MN are out of reach (WI due to massive fraud))….

  • biblerox

    LOL at this prediction model. Don’t quit your day job.
    Penn. Iowa, Wis. Nevada, and Ohio are not going to be Repubs. This will be a CLOSE race and it will NOT be a blow out by either party.
    But, nice try though. SMH.

  • lock12j

    The thing to recognize here is the change in hard party ID numbers. If the
    Gallup survey is accurate, it basically invalidates all of the polling data
    based on the 2008 numbers, and it means all of the national polls, and to an
    unknown extant at this point, state polls are building in 3-5 points for Obama
    that simply don’t exist.

    This would explain the recent Romney surge in Gallup’s 7-day tracking poll, and the
    fact that this jump was slow to materialize elsewhere (although it has, just to
    a muted extent) in the other national polls, as well as the President’s approval
    numbers running higher than his support in the general election would seem to
    substantiate.

    Rasmussen and others assemble polling results using turnout models based on the
    2008 exit polling party ID results. Gallop does not use a turnout model (they
    do, but its part of their LV screening, not in putting together final results). So if Ramussen publishes a poll of 1500 likely voters, about 600 of those voters are self-described democrats, and 450 each are either self-described republicans or independents. (just going off the top of my head after reading their most recent poll)

    Again, Gallup says they do not ask voters to state party affiliation when conducting general election polls, but they changed their methodology prior to the 2008 election
    cycle in response to accusations they were under-sampling minorities. They basically split their 7-day, 3000 LV tracking poll, or more accurately started a second 1500 LV tracking to mirror the existing one. That mirror poll disproportionately targets minority respondents until they reach certain thresholds, at which point they randomly sample respondents as needed to reach the 1500 total. It clearly favors democrats, who rely more heavily on minority voters.

    That’s why the Gallup surge for Romney in early October was so puzzling. One would have expected a shift towards Romney to first materialize in the Rasmussen poll, long accused of using voter screens to cherry-pick republican voters, an accusation I would add that rings hollow, given the fact that Rasmussen, almost without exception,
    produces the most accurate final election polling data.

  • tngal

    So don’t move to VA. Move to somewhere in Iowa or Ohio. Those two states are always in need of more Republicans and you could push us over the line. Thanks for the sWINgometer as always Neil. I know it doesn’t always come up in our favor, but its always a welcome read.

  • http://www.hakubi.us/ Neil Stevens

    Heh

    a) I moved in April

    b) I moved where the job was

    c) Virginia’s weather gets cold enough, thank you very much :)

  • http://www.hakubi.us/ Neil Stevens

    Also thanks for the feedback :)

  • http://www.hakubi.us/ Neil Stevens

    Also thanks everyone for the feedback and letting me know where the links and mentions were. Thanks very much! It was fun yesterday evening, hearing and seeing all this on the way home.

  • http://www.hakubi.us/ Neil Stevens

    He’s a now-banned troll. Here to misinform and back the President.

  • http://www.hakubi.us/ Neil Stevens

    Try reading less Nate Silver/New York Times partisan hackery and you’ll feel better. :)

  • tngal

    Ah wanderlust. I understand it well. Just know that we appreciate your efforts when it comes to numbers and polls and such. I’m so pathetic with numbers, can’t even remember my zipcode and I’ve lived here for well over 25 years. But then I take into account I don’t send myself letters very often so why do I need a stinkin’ zipcode anyway.

  • http://www.hakubi.us/ Neil Stevens

    Heh, it was the opposite of wanderlust. If I could have I’d have stayed in California my whole life. But I reached a point, or rather the state electorate and government reached a point, where I had no future there.

    I didn’t leave California, California left me. So I went someplace decent where I could get a job, and I expect to stay a while. :)

  • westcoastpatriette

    I know exactly what you mean, Neil. California has left me, too, and I’m not sure we’ll ever get it back. For now I have no choice but to stay but when the time is right, I will leave, too. Very sad. I love the Cali that left.

  • gunnyg2002

    Neil,

    I transferred from VA due to spendig about 25 hours a WEEK in a car, my POV and my govvie. I was ready for a stroke or heart attack.

    Be careful.

  • gunnyg2002

    I left California in 1987 when I left Camp Pendleton and NEVER looked back. The state is lost brother.

  • gunnyg2002

    I simply want to see Romney CRUSH Obama in a Reaganesque sort of way.

  • kirkaiya

    Looking at the two maps from swing-o-meter, I’m reminded of a common way of doing mathematical proofs back in high school and college – assume the opposite of whatever you’re trying to prove, and then when you reach a conclusion that’s absurd, you know your original premise or assumption is wrong.

    Well, when your “low end” model shows Romney winning both Minnesota and Pennsylvania, I think there’s an assumption wrong somewhere. Those two states have 30 electoral votes between them, and they’re almost without any doubt going to be won by Obama. I think the upper bound on a Romney win is probably 285, unless they win Wisconsin (which seems more doable than Minnesota – Paul Ryan’s home state is Wisconsin). You wrote that the Swing-o-meter is meant to deal with a two-party vote, and not raw party ID, so maybe that’s the problem, I don’t know.

  • http://www.hakubi.us/ Neil Stevens

    Oh look, another Obama worshipper making his first comment to dismiss this as inconvenient.

  • Common_Cents

    Neil, tell Andrea Tantaros I’m single!!!

  • http://www.hakubi.us/ Neil Stevens

    You can tell the banned idiot didn’t even read my post, where I repeatedly disclaimed a blowout.

  • Pingback: Gallup: ’2012 U.S. Electorate Looks Like 2008′ — Except That Party Affiliation Is GOP +3 Instead of Dems +12

  • Pingback: Handy orten

  • Pingback: Gallup: ‘2012 U.S. Electorate Looks Like 2008′ — Except That Party Affiliation Is GOP +3 Instead of Dems +12 :Natural Diabetes Treatment

  • Pingback: ROMNEY LEADS IN OHIO! | Union County Republican Central Committee

  • Pingback: Eden Borromeo

  • Pingback: Gregory Antill

  • Pingback: Consuelo Mastoris

  • Pingback: Lloyd Kocaj

  • Pingback: Gretta Ciresi

  • Pingback: Jamey Beller

  • Pingback: The Polls Are Wrong: Electorate is R+6 – America In Ruin

  • Pingback: Nikita Difiore

  • Pingback: Refugio Latif

  • Pingback: Chang Salge

  • Pingback: Gertie Kulesza

  • Pingback: Otha Papich

  • Pingback: Jimmie Corra

  • Pingback: Odell Sanfiorenzo

  • Pingback: Marlana Dreiss

  • Pingback: http://jcow.in/article.php?id=312

  • Pingback: Irwin Cucufate

  • Pingback: Jacquelynn Winterberg

  • Pingback: Nathan Eitzen

  • Pingback: Maureen Sabo

  • Pingback: Augustus Standfield

  • Pingback: Chance Frausto

  • Pingback: Enoch Mcklveen

  • Pingback: Mavis Wayns

  • Pingback: Ashly Saysithideth

  • Pingback: Raymond Hein

  • Pingback: Trudy Tieng

  • Pingback: Allegra Schoenfelder

  • Pingback: Sid Mellinger

  • Pingback: Marty Nobis

  • Pingback: Ignacio Godeaux

  • Pingback: Kasey Mushtaq

  • Pingback: Derek Iozzi

  • Pingback: Kylie Waffenschmidt

  • Pingback: Chad Pehowich

  • Pingback: Antonio Lishman

  • Pingback: Shaniqua Steves

  • Pingback: Clyde Honse

  • Pingback: Jim Dant

  • Pingback: Brooke Kahao