It seems there is now statistical evidence for vote fraud in the US presidential elections of November 2004. In this evaluation a group of mainly statisticians make a rather compelling case for vote fraud. The authors explain in detail that the discrepancy of the exit polls in the last presidential election with the final results is most likely due to a deviation of the official results, rather than the exits polls, from the voters' intention. The exit polls had predicted a 3% Kerry win. Such discrepancies between exit polls and final results were the primary evidence for vote fraud in the Ukraine only weeks later. In the case of the Ukraine, the US government cited the exit poll discrepancy as their motivation to back the motion for re-election. In the US, the government blamed the exit polls as false.
Source: UScountvotes