Monday, November 5, 2012

STOP THE PRESSES!

I'm stunned.  I really am not sure what to make of these numbers.  If you thought Rasmussen's September poll of partisan ID was beyond belief, well the October numbers are incredible.  To refresh your memory, this poll is conducted monthly and asks 15,000 people what party they are affiliated with.  In my Rasmussen Party ID model, I use a 3 month average of these numbers.

The new October number is way beyond anything I was expecting.  The new numbers are:

Democrat 33.3%
Republican 39.1%
Independent 27.5%

The already amazingly good September number of R+2.6 went to R+5.8!

With the new October number, the 3 month average changes:

Old  -  New
33.8    33.6   Democrat
36.4    37.8   Republican
29.8    28.6   Independent

I will be now using an R+4.2 sample for my Rasmussen Party ID model.  Expect to see that number move much higher.  My analysis below changes to Romney leading by 6.9% in that model.

53 comments:

  1. Astronautical number! If the election even gets half of your average (R+2), this thing could actually be over Tuesday night (depending on how the BG states roll).

    ReplyDelete
  2. Wow, the historical shifts of this number tell a story.

    Note that the 2010 and 2004 results followed months of -2.9 and -1.5, which (both) followed quarters of -2.1.

    We've been in the land of + pretty much ever since that 2010 election, but 5.8 is amazing. That's about *8 points better* than the conditions for the last two GOP victories!

    Can you hear that land sliding?

    ReplyDelete
  3. Dave

    Can you explain why Rasmussen is saying that the election is going to be very very close, if his own party ID numbers point to a blowout?

    NJH

    ReplyDelete
    Replies
    1. If you make two predictions, you double your chances of being right

      Rove did it too

      Delete
    2. Because he doesn't believe his party ID numbers, or doesn't believe turnout will match them. He is invested in his tracking poll and unwilling or unable to publicly acknowledge that it is flawed.

      Delete
    3. and Dave's good at it too, he has 6 ! (wink,grin)

      Delete
    4. I will dial that back to 3, though ;)

      Delete
    5. It's a hedge I think.

      If he says Romney will win 52-46 in PV and it ends up being 49.8 to 49.3 for either side he will look like a fool.

      But if he says it's 49/49 Romney and Romney wins 52-46 he'll say well I predicted the right winner, I just missed out on the unprecedented Republican surge.

      And since everyone else is predicting an Obama win, that prediction will still be better than being so wrong the other way.

      Which is reason # 271,288 why polls don't mean a hell of a hot anymore. Polls are more political than politics.

      Delete
    6. I am in need of Rove right now. I find reading his face as he gets out his white board tells me a lot about how he's feeling about OH.

      Delete
  4. I don't even know what to say about this.

    I mean, holy wow. Is this possible???

    ReplyDelete
    Replies
    1. I don't know.

      The number is so good that I find it hard to believe.

      Delete
    2. Bronco Bammy made people embarrassed to be a Democrat with those silly debate performances, and Romney looked like leader people wanted to identify with. D's had a lot of very bad days in October.

      Delete
  5. Ras party id poll has been spot-on in 04 and 08

    http://battlegroundwatch.com/2012/10/08/will-the-party-id-really-look-like-r-2-or-r-3/

    ReplyDelete
    Replies
    1. Interesting... Though using 3 (as Dave suggested) seems to be safer.

      Delete
  6. Don't the EV results point towards the Rass numbers as being closer to truth.

    ReplyDelete
    Replies
    1. Definitely, EV is down in D counties from 2008, and up in R counties

      Lets see if Nate Silver flinches today.

      Delete
  7. These state polls are just crazy. Here's the CNN Ohio poll that says election day voters are favoring Romney 55-42, this after the data from Board of Elections say Rs have reduced the early vote advantage by more than 200,000 votes over 2008.

    ReplyDelete
    Replies
    1. CNn poll link

      http://politicalticker.blogs.cnn.com/2012/11/02/cnn-poll-obama-50-romney-47-in-ohio/

      Delete
    2. Not going to trust any poll from CNN after their D+11 distribution for the national one.

      Delete
  8. The fact that the house will stay GOP - and maybe even go more GOP - means a 2008 electorate is impossible. Does anyone (aside from Nate Silver that is) believe the same 2008 electorate that have Dems an 80 seat win in 2008 will RE-ELECT Obama **AND** give the GOP a 50 seat win in 2012? It defies all sorts of logic to believe this.

    ReplyDelete
  9. So the Gallup poll has been released (well, unofficially I guess). 49-48 Romney.

    ReplyDelete
  10. The Lord givith (Ras Party ID) and he taketh away (Gallup).

    In all seriousness, I am feeling better about tomorrow looking at the EV numbers (esp. in Colorado) and how the Dems are down and the Reps are up.

    ReplyDelete
  11. Oh please. Gallup went from a weeklong tracker (I think) to a pure weekend sample. And we don't know the internals.

    ReplyDelete
  12. great news, but HotAir is now reporting that the leaked last national Gallup is showing Mitt up by only 1 and that's an erosion of his previous 5. ACHHHHHHHHH. Just when I am really convinced we are going to win, I see THAT!!!!

    Now, admittedly, I haven't even read your post yet, just your line that the Rasmussen is good news, so now I'll go back and really read it.

    Thanks.

    ReplyDelete
  13. The problem is that when a major pollster like Gallup gives Dems hope, they may be more energized to show up.

    ReplyDelete
  14. Wouldn't you think that after the early October first debate in which Romney killed Obama, many people called said they identified as Republican? You know, the old, "I'll go with the winner" thing?
    However, as the debate season wound down and debates 2 and 3 were seen as ties or spun as winners for Barry, that mind set probably disappeared, as reflected in the erosion in Gallup for Mitt.

    ReplyDelete
    Replies
    1. IDK. I remember seeing the tracking polls days after the other two debates and despite the media spinning, the numbers didn't move that much and Romney's momentum continued on.

      Delete
    2. Yeah, that's true. So, is this nothing more than storm related, media spinning for Barry that is working, and the usual return to party at the end of a campaign?

      Delete
    3. Look, the whole reason why we look at party ID in the first place is that it's more stable. A switch isn't just "this policy sucks", it's closer to opening your eyes and realizing how crappy (or, from I->partisan, acceptable) one side was all along.

      Once you've had an experience like that, you aren't suddenly going to swap back a few weeks later.

      Delete
  15. I try not to take Dick Morris too seriously, but these Party ID polls from Ras and Gallup have shifted 10-15 points and have a decent track record of being correct. How does that mean nothing?

    As for Gallup, that's every poll now breaking towards the middle. I struggle to believe any of them.

    ReplyDelete
  16. Gallup is based on the sandy hits days. Left out the aftermath days. No worries.

    ReplyDelete
    Replies
    1. Are you sure. I read on HotAir it was a weekend only sample. I guess we have to wait for Dave to look at the internals.

      Delete
  17. Can you explain the discrepancy between Ras' early voting numbers and the EV numbers that are reported from the registrar of voters in Ohio?
    From Hot Air:

    " The bad news for Romney? According to Rasmussen’s poll, Obama has a 23-point lead among those who have already voted. Those numbers don’t match up to the early-voting numbers shown by Ohio’s official reports, however, where Republicans have gained over 260,000 in the gap over 2008. Regardless, Election Day voters are much more Republican; 71% of Republicans will go to the booth tomorrow, while slightly over half of all Democrats have already voted. Among independents, 40% have voted early.

    "In Rasmussen’s poll, that may also be bad news for Romney, depending on which 60% of independents remain. He trails among unaffiliated voters in this poll by eight, 50/42, although 17% say they could still change their minds. However, the gender gap has been completely neutralized, with 52/45 splits among men and women breaking in opposite directions. Obama has a favorable job approval rating, 51/48, in an R+2 sample (37/39/24)."

    That last line really makes me wonder about any of these polls. HOw the hell can Obama have a fav job poll rating with an R+2 when almost all year he had less than fav from national polls?

    ReplyDelete
    Replies
    1. If the job approval rating of 51-48 is accurate, that's not good.

      Delete
    2. Dave already did. Dems are lying to pass the voter screen.

      Delete
    3. If they "are lying to pass the voter screen" in places like Ohio, it's because Obama has a ground game that's been in place forever to train them to do that to scew the polls. Why else would a regular, non-activist Dem lie to do that unless the party had trained them to?

      Delete
    4. Well, that was Dave's explanation. Mine is slightly different -- I think people are more inclined to answer pollsters once they've voted, esp. if they've voted the "right" (social pressure) way. Remember, the response rate on these things is *terrible*. I hang up on every poll myself.

      Delete
  18. There doesn't seem to be much of a reason to believe any of the polls at this point.

    ReplyDelete
    Replies
    1. Indeed.

      While the headline results are similar, polls vary wildly on independents, early voters, how early voters are voting, etc... to the point where you can't take any of them seriously.

      It's going to come down to turnout. D+3 or better, and I think Romney's got this.

      Delete
    2. This is where I'm at right now in my thinking. Either Romney wins this - by some amount beyond the "margin of fraud" - or else the country that I'm working and hoping to save by electing Romney no longer really exists except in my mind. I don't see how else all the evidence *except* for the polls can be wrong and the polls right, unless that last part is true. If someone like Obama with his record can't be voted out of office, then we're already living in just another debt-ridden Eurozone nation; all that's left is to go through the rest of stages of grief. I don't think we're that far gone yet. I fervently hope that we're not that far gone yet, but we'll only find out for sure sometime late tomorrow or early Wednesday. It's for sure these pieces of propaganda masquerading as polls aren't going to tell us before then. Just got to go vote, keep our fingers crossed, and pray hard if you go in for that sort of thing.

      Delete
  19. LOL - Gallup apparently includes Thursday (Chris Christie fellatio day).

    O is toast.

    ReplyDelete
  20. From Rasmussen's VA poll (via Hot Air):

    "Mitt Romney still earns 50% support in Virginia just before Election Day.

    "The latest Rasmussen Reports telephone survey of Likely Virginia Voters shows Romney with 50% of the vote to President Obama’s 48%. One percent (1%) likes another candidate, and another one percent (1%) is undecided.

    "This is unchanged from two weeks ago and the week before that when it was Romney 50%, Obama 47%.

    "This one’s tight enough to look at the internals, which are somewhat surprising given the closeness of the toplines. Obama actually loses the overall gender gap by three points (-7 among men, +4 among women), but he’s also losing independents in Virginia by 21 points, 58/37. In 2008, Obama had a +11 in the gender gap and won independents by one point, 49/48. The D/R/I in this sample is D+2 at 38/36/25; in 2008 it was 39/33/27 but in 2009′s gubernatorial election it was 33/37/30.

    "Romney wins the economic argument by six points, 51/45 over Obama. There’s a significant gender gap on this question as well, but it also favors Romney (+10 among men, +1 among women). Romney has a 25-point lead among independents on this question, 58/33. On the other hand, Obama does have a positive job-approval rating at 51/49, which is probably why the toplines look as close as they do. I’d guess, though, that Virginia’s going to break significantly for Romney."
    ____________________________________________________________________

    This is what I don't get: in spite of the internals, "Obama does have positive job approval rating at 51/49."

    So, does this essentially mean that a voter giving him a positive approval is thinking, "Well, he's okay, but I think the other guy can do better"?

    I mean, if you're going on record to say you are voting for the other guy I(Romney's at 50%, 2% over Barry, then you'd think those 50% would give Barry a negative job rating.)

    Maybe it's just the margin of error stuff, but I'm really confused. Why would you vote for the other guy if you think this guy is doing a positive job? That's what's scary. Mabye the word is just too loaded, with too many meanings, and all that matters to certain voters is single issues rather than "overall" ratings?

    ReplyDelete
    Replies
    1. the job approval numbers are inflated by Sandy. People are being polite to their president.

      Delete
  21. I don't think the electorate is going to have an R+6 advantage, you would see something like a 45 state landslide if you took that number and also added independents backing Romney by double digits.

    But I don't think anyone can say with a straight face that Democrats are going to be able to do much better than pull about even with Republicans in turnout, in which case, almost all the polls are wrong by a huge margin. If CNN's "numbers" are correct, but you adjust for an even turnout, it's a Romney Rout.

    My bet is something like R+1. I just hope it's enough to drag some of our "less than optimal" Senate candidates across the finish line.

    ReplyDelete
  22. This comment has been removed by the author.

    ReplyDelete
  23. Dave,
    After this election you need to find a new hobby. ;)

    ReplyDelete
  24. Saw a comment on a website that said that party ID doesn't mean anything during an election year where only one party had a primary contest. This is because many people switch party ID's to be able to vote in them, often for malicious purposes.

    What do you make of this?

    ReplyDelete
    Replies
    1. That would only make sense if this happens during that primary, but here the party ID (in Ras) has increased in October, far after the primary was over.

      Delete
    2. I think that line of logic confuses party identification with party registration. I was under the impression that Rasmussen and Gallup do "party id" surveys rather than just gather party registration data or party id data from primaries.

      Rasmussen releases numbers on it every month and Gallup weekly or every ten days, something like that. I doubt it's just numbers from actual registration but I could be wrong.

      Delete
    3. I think it does, and it also means that the red shift is still happening

      look what's happened since Obama got elected

      Chris Christie
      Bob McDonnell
      Scott Brown
      Weiner's seat
      2010 Midterms
      Scott Walker recall

      This means it's still happening

      Delete