Reply #8: Of cours it's rather difficult to evaluate an abstract [View All]

My understanding is that the ESI study was much like Walter Mebane's study which was included in Section VI of the DNC report. Several strong correlations were shown between the Kerry vote and various other variables (such as Fingerhut vote, Issue 1, Gore 2000 vote) by precinct, and from that it was concluded that the amount of fraud could not have been great enough to switch the election results.

I have two problems with that:

1. This is the least of the two problems that I have. Whereas it is true that strong corellations of the Kerry vote with other variables does suggest a limit to the amount of fraud that could have occurred by electronic vote switching within precincts, it is not clear to me how much fraud could of that sort could have occurred, while still maintaining those correlations. Maybe the researchers have a handle on this, and they just weren't able to explain it to us very well.

2. But the more serious concern is that, even if they were correct in their above conclusion, that does not rule out by any means fraud committed by electronically adding votes to Bush precincts or subtracting them from Kerry precincts (by means of central tabulators), in proportion to the votes of those precincts that were legally counted. Therefore, I see no basis for such a sweeping conclusion on their part.