Interesting 'Turn Out Model' research by ... me :)

[landrun]

I've heard the complaining about polls over sampling dems/repubs to get their numbers.
I'm not convinced who's right and haven't figured out how they derive that variable.

But I read where Rasmussen's party identification has been spot on in recent elections.

So, I've put together a table using both Rasmussen and Gallup's party ids to date to try to determine what a likely turn out model will be.

I've listed the year, the percentage of the voters who were dem and republican.

Then I list this years third. Finally I list 2010 - just to make a point.

In short, if the voters are consistent like they've been over the last few election cycles, this could be huge for Republicans. So huge, it is hard to believe really.

Notice this, In 2004 Repubs won - with an average of 1% more Dems than Repub - 43.5-42.5
Dems won BIG in 2008 with over an 9 percent advantage 47 dems and 37.5 repubs
This year, for maybe the first time in my life (I haven't finished looking but my perception is that this is true) the Repubs actually have the advantage by 2.5 points. 40 to 42.5 for the repubs.

How big might this be? In 2010 the Dems had a 2.5 margin and the Repub gains were historic.
How big could it be when more Americans actually identify themselves as repubs rather than dems??

In any case, this difference looks better for Repubs than any election in a long while.


<pre>
gallup ras avg
d r d r d r Difference
2004 48 48 39 37 43.5 42.5 1
2008 54 42 40 33 47 37.5 9.5
2012 46 49 34 36 40 42.5 -2.5
2010 31 29 36 33 33.5 31 2.5

</pre>


Here's the sites where I got my data...
http://www.gallup.com/poll/125450/party-affiliation-despite-gop-gains-states-remain-blue.aspx
http://www.gallup.com/poll/145463/democratic-party-drops-2010-tying-year-low.aspx
http://www.rasmussenreports.com/public_content/archive/mood_of_america_archive/partisan_trends/summary_of_party_affiliation

[landrun]

Well, conservatives are starting to hint that this might be a blow out for Romney:

http://www.breitbart.com/Big-Government/2012/10/29/Gallup-Shock-Romney-Up-7-with-early-voters



**GALLUP SHOCK** Romney Up 52-45% Among Early Voters

But here is what we do know: 220,000 fewer Democrats have voted early in Ohio compared with 2008. And 30,000 more Republicans have cast their ballots compared with four years ago. That is a 250,000-vote net increase for a state Obama won by 260,000 votes in 2008.


I know this is all just guess work. But a week out, and I feel like Romney actually has a chance to win this thing.

[hawaii 5-0]

Breitbart is right up there with Rasmussen as far as being biased toward one particular persuasion.


Whatever. I'm still gonna survive one way or the other.

5-0

[landrun]

Breitbar is conservative. But all they did was point out facts from Gallup's polls (not Rasmussen)

Rasmussen has been ranked in the top two as far as accuracy goes over the last two election cycles. He appears more 'Republican' only because he consistently uses a likely voter model rather than registered voter. Most other firms use register voters until the last month or so of the election and then they start using likely voter also. And when they do, they generally move to the Republicans a little.

I thought the exact same thing and dismissed Rasmussen for a while because I thought he was 'bias'. But after doing some research on the polls I discovered the reason his numbers look the way they do, how accurate he was and how in the end, the other polling firms use likely voters too. If you doubt this, just go to realclearpolitics.com and look at their poll of polls. Today, almost everyone is using likely voters. A month ago or so, most were using registered voters.

[XingTheRubicon]

Breitbart is right up there with Rasmussen as far as being biased toward one particular persuasion.


Whatever. I'm still gonna survive one way or the other.

5-0


Rasmussen was the most accurate poll in 2008. Have a nice day.