Sunday, November 4, 2012

November 4, 2012 Analysis


Polls included:

Tracking Polls
R+0.8 - Rasmussen Daily Track
R+5 - Gallup
O+1.3 - IBD/Tipp
Even - ABC/WaPo

Others
O+1 - Battleground
Even - Pew
R+1 - NPR
O+1 - CBS/NYT
Even - Fox
O+5 - National Journal

O+0.15% - Current RCP Average
O+0.73% - Average using the 2008 turnout model
R+2.05% - Average using the D+3 turnout model
R+4.28% - Average using the 2010 turnout model
R+4.33% - Average using the 2004 turnout model
R+6.06% - Average using the Rasmussen Party ID turnout model

Average Democrat advantage across all polls: D+4.16

The national polls continue to show a tie.  They also continue to assume a Democrat turnout advantage of over 4 points.  I think we can safely say that if the Democrats are able to turnout 5,200,000 more voters than Republicans, then the race will indeed be close.  As I said yesterday, I view such a turnout advantage to be very unlikely.  I believe that as we moved closer to the election, too many Obama voters were making it through the likely voter screens by falsely claiming to have already voted.  This was especially prevalent in state polls, but it also is affecting national polls.

The only polls that released new numbers were Rasmussen and ABC/WaPo.  Rasmussen showed no change, while Independent support for Romney returned to +4.  ABC/WaPo dropped into a tie.  This was because their sample expanded back to D+4.  Oddly, they also found that Independents have no preference, with equal support for Obama and Romney, an 8 point swing in one day.  I believe this result to be unlikely, but I included their result in the averages.

Overall, my conclusions from yesterday remain the same.  If turnout is D+3 or better for the GOP, then Romney will win the election.

Tomorrow everyone will be releasing final polls.  I intend to post analysis of their results, and make a final prediction.

21 comments:

  1. That ABC/WSJ poll, esp with Independents, is so odd. Wasn't it +4 for Romney, then +8 and now even? Apparently the Independent voters are indeed fickle! :p

    ReplyDelete
    Replies
    1. One of the reasons I'm suspicious of what is being reported. Voters aren't fickle. They decide and don't change their minds.

      Delete
    2. Yeah, I know.

      If Romney does go on to win, I wonder how the pollsters will adjust their methods for 2014 and 2016, because something has changed (I'm talking about the legit ones).

      Delete
  2. The ABC poll looks even more suspicious in light of the new polling in Pennsylvania, Minnesota and Michigan with Romney tied or slightly ahead. Weren't these state level polls what Nate Silver and his followers were depending on to save Obama?

    ReplyDelete
    Replies
    1. Yes, but keep in mind the ABC poll is D+4. A tie in a D+4 electorate will certainly lead to tied races in states that tend to be about D+4.

      Thing about the PA poll is it is D+6 and showing a tie.

      Delete
  3. Even without reweighting, feeling a bit more confident after comparing mean standing to this day in 2004.

    Bush was at 49% and leading by a point in a half.
    Obama is knotted-up with Romney @ 47%

    Republicans seem to have a slight shy tory factor, which would net a slight advantage to the R in the race. Bush improved by around 1.7%, Kerry by around 1%.

    That's using the RCP average.

    ReplyDelete
    Replies
    1. Don't forget that Kerry then went and won the Independents by 1 point. Romney holds that edge now. I'm seeing at least a 6 point edge, and I think it is under reported.

      Delete
  4. Not so much to do with these sets of polls, but it still puzzles me why so many pollsters believe there will be such a high turnout amongst Dems. They must know that Obama's support won't be the same as '08 AND the Reps have much more support for Romney than McCain had.

    ReplyDelete
  5. Maybe a combination of lower response rates and lower poll budgets. That results in lazy likely voter screens.

    ReplyDelete
  6. I hate to be an eeyore, but this is what many of my pessimistic anarcho-capitalist chums think:

    If these polls with high D+ samples are correct with the job Obama and a mostly rubber stamp D congress has done, forget swing states. The United States would be a solid D country. I don't see how the GOP could ever have a President again.

    ReplyDelete
    Replies
    1. Well, if we really do have a D+4 or D+5 result, then they are correct. We have irrevocably become France. Maybe we can get topless beaches.

      But Rasmussen, Gallup, and now Pew are saying the party identification right now is between Even to R+3.

      Delete
    2. Forget who turns out and what the turnout is in the future. If Obama beats Romney, we have lost the republic.

      So yeah. You don't need to be any sort of anarchist. Just get out and vote.

      Delete
  7. I'm going with a big win on Tuesday.

    Check out this site reporting absentee voting in Virginia.

    Look at Fairfax County that went 60% Obama, they are way off their 2008 numbers.

    http://www.vpap.org/elections/maps_absentee

    ReplyDelete
    Replies
    1. Pay close attention to Loudoun County. That was prime territory for Obama in 2008, but it is Romney country now. Being equal to 2008 is actually a big shift to Romney.

      Delete
  8. When I surf around these election reporting sites, in many more cases R counties are up and D counties are down. That's movement in two directions.

    ReplyDelete
  9. Dave,
    Do you think a Dem cross over vote will come into play or is that figured into one of the models?

    My anecdotal evidence, is my liberal leaning daughter in CA will not support Obama this time, my in-laws in KS (think teachers union) already voted Romney. Friends I talk to in MO seem to be Romney supporters (BTW, will probably vote Akin as the lesser of the evils). I'm having problems talking to anyone publicly supporting Obama.

    I don't know if it means anything or is this a small piece of a larger picture. I'm on the border of KS/MO. Your thoughts?

    ReplyDelete
    Replies
    1. I actually discount cross over vote in my models. Obama will not win a cross over differential, but if Romney does then my models are a bit pessimistic.

      Delete
    2. My dem friend in CA voted Romney.

      Delete
  10. I am teetering back and forth - so am trying not to think about it. However, what I can do that helps me get some sleep is ask myself do I know "anyone" who crossed over in 08 due to weariness of the Bush years and voted for Obama - I do. Then I ask myself do I think "any" of the Republicans and the high amount of Evangelicals who sat out or crossed over will stay in President Obama's camp. I don't think they will - then again I'll be going through the next 60 hours as a human ping pong ball!

    ReplyDelete
  11. Dave, have you seen this:

    https://docs.google.com/spreadsheet/ccc?key=0AvEOdIaw0fPNdGtuV3pidFpWU3B3cDBtLWR4cGdZRmc#gid=0

    and this

    http://boe.cuyahogacounty.us/pdf_boe/en-US/2012/2008_2012InHouseVotingDailyComparison.pdf

    and this

    http://www.vpap.org/elections/maps_absentee

    I think what we are seeing is an electorate that will be more like 2004 than 2008

    ReplyDelete
    Replies
    1. It used to be said nothing fires up the GOP base than Hillary Clinton. I think that she can't light a candle to President Obama on that count. Republicans have been salivating since '10 for this election.

      Delete