More spin from the corporate media:
ABC news just commented on the disparity between the exit polls and the overall outcome of the election. They're saying there were problems with the polls, that they oversampled women, oversampled early voters, etc., and that Democrats are usually more willing to be interviewed.

This is IMO a feeble attempt at a whitewash.
First, exit polls have a long history of being extremely accurate in predicting election results, and the poll results cited at Randi Rhodes's site were not just the early poll results.
(The TV news has been dangling poll results in front of the viewers for MONTHS and analyzing them as though they are assumed to be done properly. NOW, when they don't like what seems to be their meaning, they're all of a sudden suspicious of their validity! Exit polls are MORE useful than pre-election polls, because everyone polled already HAS VOTED.)

More importantly, -- HELLO! -- it's not a question of how accurate the polls are. The question is
Why were the polls more accurate when paper ballots were used than when people voted via electronic machines that had no paper trail or other means of doing a recount?

Geez, Jennings!

Genie