What are the exit poll percentages for the millions of people who sat out the last election as well as the 2008 election?
I'm sure that if the sort of polling numbers you place so much value in are actually informative rather than deceptive you'll have the figures for who didn't vote right at your fingertips.
How about the total number of voters in Ohio who didn't vote because they refused to vote for a Mormon along with the way they voted in prior elections?
Surely you must have that sort of number at least by State if not finer granularity, or do you just stick to national numbers in order to avoid reality which proves that Zip code is far more predictive of how people vote than any of the garbage breakdowns by religious group? Especially breakdowns that count people in various groups who haven't even been to church in the past six months or more.
See post 41.