10/08/2012

Today's Polls 10/8/2012

Via-Jer's Notes



Most of the excitement today has been about the new Pew Poll which shows Romney suddenly leading 49 to 45. I have never had much use for the Pew polls so although I would like to believe it is true, I don't really trust it. I suspect it is a set up to show later in the month that Obama has regained momentum. I have found Pew over the years to be the most biased of all polling companies and I doubt they have changed their stripes.

The other major poll today I find much more interesting even though it still shows Obama ahead is:



Politico/GWU Battleground Poll

National

Result
Obama 49% Romney 48%

Turnout Model D/R/I (includes leaners)
D+4 (44/40/16)

Romney keys
Independents 51-35
Suburbia 55-43
Whites 56-41
Hispanics 38-60

Obama's Job approval 
in Florida  38%
Suburbia 44%
Independents 35% (48% strongly disprove)

Notes of Interest

Even though, in my opinion, they slightly over sampled Democrats the more telling distortion was that they over sampled those who call themselves "strong" Democrats (30%) to those who call themselves "strong" Republican (23%). Although this probably does not make much a difference in the "horse race" between Obama and Romney it skews the issue questions. Even at that the Republicans were more enthusiastic to vote than the the Democrats 86%-75%. Also and far more significant, 85% of this poll was conducted prior to the debate!

Note too Romney's keys. there is absolutely no way Obama could win an election losing the Independents by 16%. I would say though that the way they divided up the Independent "leaners" there is no way to be sure that this is a true representation of Independents. Since the 16% represent only those that self identify as those who do not lean towards one party or the other you need to add back in the "leaners"  to see how they would vote as a block. But indications are that it would be even stronger for Romney since 10% lean Republican and only 6% lean Democrat. If you put the "leaners" back into the Independent column the actual D/R/I would be 38/30/32 so in fact the actual slightly oversampling of Democrats is more than it is made to appear.

So this poll which only shows Obama up by 1% which was conducted  mostly before the debate, over samples "strong" Democrats compared to "strong" Republicans in a nation in which conservatives out number liberals two to one. Additionally by oversampling Independents then dividing them up and "making them" Republicans and Democrats they are able to show a D+4 sample instead of the D+8 poll it truly is. Clever. But even with all of that the best you get is Obama by 1%!

The reason for this is obvious, no matter what they do with the party affiliation numbers they can not overcome the very important fact that this poll shows that Obama is not popular in the suburbs and is losing there badly. Considering he won the suburbs by 2% in 2008 if he actually is loosing the "burbs" to Romney by 10% that is huge. Just from a demographic stand point that would put states like Pennsylvania in play (which I believe it is) because the suburbs of Philadelphia went for Obama big in 2008. It will also have a big affect in areas like the I-4 corridor of central Florida which is always a key in Florida elections.

Finally if Romney is anywhere close to getting 38% of the Hispanic vote, election over. I doubt he is, this probably is based on too small a sample, but still.

I don't see anything in this pol that is good for Obama or bad for Romney, we will have to see how things unfold but I believe there is the beginning of a Romney surge.


No comments:

Post a Comment