[kesg]

I fully agree with everything you said. To which I add that I think that Gallup is allowing early voting numbers to skew their model. They are assuming that because blacks are voting in disproportionately large numbers in early voting, they will continue to vote in those same disproportionately large numbers on Election Day. This is simply wrong. What really happens is that blacks who vote early don’t vote on election day. They still make up only 11-12 of national voters, regardless of the day on which they vote. Well, except perhaps in Chicago....but I digress.

There is no basis for Gallup to assume that because more blacks are voting early, the total population of black voters will increase as well. Yet that is what its latest LV model does.

[goldstategop]

Good point. If blacks have already early voted in states where early voting is allowed, they don't vote again. The assumption is because of Obama's race, we'll get an an exception to the normal pattern with regard to Democratic candidates. But even if that is true, that does not change the ratio of black voters relative to the population. Thus, even their numbers increase, you also have an increase in white voters, which cancels out the effect of the increase in black turnout for Obama. And even if 9 out of 10 black votes do go to Obama, piling on more votes does nothing to move the overall popular vote numbers to where Gallup thinks it be will be. John Kerry increased his voting by the millions but he didn't increase the percentage to a winning margin. Its highly unlikely that Obama can do what Kerry failed to do four years ago even if he does get millions of new voters to vote for him on Tuesday.

"Show me just what Mohammed brought that was new, and there you will find things only evil and inhuman, such as his command to spread by the sword the faith he preached." - Manuel II Palelologus