POLLING: THE MORE WE UNDERSTAND, THE LESS WE KNOW

Rasmussen and Gallup have the national race at Obama +2, within the "Margin of Error" (I hate that term, but that's another whole post.).

Yet when you peel back the layers, there is no cross-tab or demographic that supports McCain optimism, even veterans. Going State by State, Johnny Mack trails by "Outside the Margin" amounts in high Electoral Vote States and trails in traditionally safe GOP States. Why the disharmony between the National Polls and the individual State by State Polls?

They are not sharing the same assumptions about who is going to turn out. There are two issues which cloud this badly: First time registrants and cell-only contacts. The two overlap but are separate concerns.

On first time registrants, most pollsters use a factor to decrease their likelihood of voting. Every pollster uses a different factor. Are they highly motivated or just signed up at the union hall or PTA meeting to get the deputy registrar out of their face? Unknowable in this "Change" year.

Cell phones are still indexed badly so finding cell phone users in order to poll them is tough task. Secondly, a lot of them are not property owners. People who don't own real estate tend to turn out in lesser percentages. Nobody really knows about Cell Only people. They are a new breed.

As a political worker, if I could have one piece of data arising from this election (solid, after election info) it would the rate @ which cell only citizens, already registered, actually vote. I thinking it will be about the same as college students, generally low. But data would be much better than guessing.