Conservative and Liberal Pollsters: Obama by a Landslide
Michael Collins at Scoop USA today posted a fascinating article entitled, "Votemaster, TruthIsAll, & Election * Projection Call it for Obama - Big!"
The results calling for Obama to win the electoral vote by as much as 368-170 and for the Illinois senator to win the popular vote by 7-9 points is huge considering the disparity in the 2004 data between these three independent polling analysts.
These polls are especially valuable when one considers that Votemaster is libertarian, Election * Projection is very conservative while TruthIsAll is unabashedly liberal. Yet these three polling analysts came up with virtually identical numbers despite representing a wide cross-section of political bias/ideology. These results are also very notable, to say the least, because all three places used a more comprehensive methodology and a running mathematical model, taking a nation-wide poll.
Polls done by places like AP/Ipsos, CNN, Zobgy, Gallup and so forth vary so wildly and are therefore suspect because they're often not nationwide, because political bias does influence the results and often the language of the questions is badly-worded. Plus, crucial data and demographics are not factored in.
For instance: A month ago, the Pew Research Center did a limited poll in which landline users and/or cell phone users were asked whom they supported. The results were startling.
Pew discovered a factor previously ignored by other polling agencies: That cell phone users were more prone to vote for Obama while landline users were more or less evenly split between Obama and McCain (The September polling showed virtually a dead heat between the two candidates among strictly landline and landline/cell phone users). Since young people, who overwhelmingly favor Obama, are infamous for living and dying by their cell phones, you would think that someone before Pew last month would've factored this in.
In June 18-29, 61% of cell phone users supported Obama compared to only 32% for the technologically-challenged McCain. This revealed an important undervote for Obama that could significantly skew all the conventional polling done up to that point, considering how many millions of young cell phone users 18+ there are in America.
So read Collins' article and follow the links. If you're a political junkie like me (and why would you be here if you weren't?), it'll make for fascinating (and comforting) bedtime reading.