9/11/2012

Live by the polls, Cry by the polls

Via-Jer's Notes




I'm a poll watcher, I can't help myself, I'm addicted. The one good thing about my obsession is that I have taken a great deal of time studying polls and understanding them. Polls like most anything else are only as good as the what goes into them, garbage in/garbage out as they say.

Take the newest ABC/WAPO poll which has the headline;


Obama, Romney run close among likely voters after conventions

Then there are over sixty questions about the candidates only two of which are at all meaningful. The first meaningful question is the fourth question which asks:

Q: (Among likely voters) If the presidential election were being held today and the candidates were (Barack Obama, the Democrat) and (Mitt Romney, the Republican), for whom would you vote?

Pretty straight forward and the answer to this question is the reason for the headline, Obama 49% and Romney 48%.

Yes obviously very close. They then proceed to ask over 50 questions about the candidates and the issues...all of which are meaningless. Why are they meaningless? Because of the very last question:

Q: Generally speaking, do you usually think of yourself as a...

Now the answer to this question tells us a great deal about that "who would you vote for question, why? Because this is how they answered "Generally speaking, do you usually think of yourself as a..."

Answer Democrat 33%, Republican 23%, Independent 37%, Other 4% and Don't know 3%.

Putting aside that 3% of the respondents don't know what they are (which probably means their Democrats who don't want to admit it) this poll is actually very bad news for...Obama. As I have pointed out before in 2008 (the worst Republican turn out since 1980) the breakdown was 39% Democrat, 32% Republican and 29% Independent or a D+7 advantage.

So in order to get a result that showed Obama in the lead, not only did they have to over sample Independents, at a rate far beyond their actual numbers and Republicans far below their actual turnout even in 2008, they gave the Democrats a D+10 advantage over the Republicans. Not going to happen.

That is why all those questions on issues are meaningless, Republicans do not make up anywhere near that small a portion of the electorate so any questions with such a skewed party affiliation does not represent a true reflection of the electorate, or the country. But it is worse (for Obama) than that.

In the details of that "how would you vote" question we find these results all those extra Independents would vote for Romney over Obama by a 11% margin, 54% to 43%. This is why they had to have such a large unrealistic disparity between Democrats and Republicans because had the even have gone with a D+7 as it was in 2008, Romney would have a lead, a rather large lead. Probably worse for Obama is something else in the details of this question, despite both the skewed party affiliation and the so called Republicans "War on women" Obama only leads Romney among women by 5%, 52% to 47%. If the poll had accurate party affiliation it is likely that Romney would not be trailing Obama at all with women.

So why do two giant news organizations ABC and the Washington Post use such skewed and inaccurate polling numbers? Well they are biased beyond anything we have ever seen in America and they want Obama to win.

We can only hope that Obama's voters believe these and other unrealistic polling numbers and think they have it in the bag and don't show up to the only poll that really matters on November 6th.

No comments:

Post a Comment