Maybe the Conspiracy Theorists Are Right About the Election

So the media have released a report that the exit polls did show Kerry winning. BrooklynDodger, not having seen the report, still wonders whether the change in results was a time of day phenomenon, with white, married people voting later than others. The Times account suggests that exit pollers approached younger people rather than older, to skew the results, but the statistical model should have taken care of this.

The alternative explanation is that the exit polls - asking people how they thought they voted - are a more accurate depiction of how they wanted to vote than the votes. This was certainly the case in Florida in 2000, where even in the absense of the butterfly ballot, Kerry won the election [and the Presidency.] Next comes the question of mechanism, especially in Ohio. Was it machines and computer fraud, or did the Democrats simply vote wrong more than Republicans?

http://www.nytimes.com/2005/01/20/politics/20exit.html?oref=login