I do agree that, in general, Democrat voters answer pollsters more frequently than Republicans do. If the pollster doesn't make modifications in his raw data to account for this bias in favor of the Dems, he comes out with a bottom line that overstates the percentage of the vote that the 'Rat candidate will achieve. But this is not what happened in the 2012 presidential election in swing state after swing state. The opposite occurred: usually reliable pollsters underestimated Obama's ultimate reported percentage of the vote. In fact, state after state where Romney was said by pollsters to have led by one, two, or three points was won by Zero (at lest if you believe the reported vote counts). So I don't see how your assumption about Dems being overrepresented in answering pollsters - though probably true in most instances - relates to what actually happened in last year's presidential election.
The most logical way to explain the discrepancy between the polling results and the reported vote totals is that the reported vote totals were effected seriously in Zero's favor by fraud and cheating, factors which no pollster can estimate before the election is completed and ballots are tabulated.
It's not so much Democrats answering polls more frequently ~ it's about subgroups like the NEA, the abortion rights fanatics and the homosexuals answering polls every single time!
That's been happening all along. It's the reason why gay marriage polls show it's popular, but when put to a vote it generally loses.
The only way pollsters can deal with this is to get out of the polling business.
Actually the opposite occurred. There was a significant spit-at-home vote due to Rmoney throwing Conservatism and Tea Partiers under the bus.
And, Tokyo Rove supported him. Q. E. D.