Much anecdotal information suggests that new voter registration (prior to the 2004 election) in Democratic areas of Ohio vastly exceeded new voter registration in Repbulican areas of Ohio:
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=382415&mesg_id=382559. Indeed, this was a major reason why many believed that Kerry would win Ohio in 2004, and why many still believe that he did in fact win Ohio.
Yet, official figures show that votes cast for president in Democratic Ohio counties in 2004 exceeded those cast in 2000 by no more than the increase in presidential votes from 2000 to 2004 cast in Republican Ohio counties.
The recently released DNC report on the 2004 Ohio Election claimed that their analysis strongly suggested that there was "no widespread fraud" in that election. This was based upon the demonstration of a strong correlation between Kerry's share of the vote and some other variables, such as percent vote for Eric Fingerhut, the Democratic Senatorial candidate. Yet, this analysis did not even consider the possibility that fraud was committed by electronically eliminating votes in highly Democratic precincts or counties or adding them in Republican areas, which could have occurred without interfering with the above noted correlations.
The absence of any difference in the increase of votes cast for President in Democratic vs. Republican counties, despite the apparent massive increase in voter registration in Democratic areas of the state suggest that election fraud could have occurred in the manner noted in the above paragraph. The DNC report finding that voter turnout in 2004 increased in precincts where new voter registration was up further adds weight to this possibility.
If it can be shown by actual data that the reports of massive increases in voter registration in Democratic areas of the state (compared to Republican areas) did indeed occur, that data could be used to show the extreme implausibility of a Bush victory in Ohio in 2004, by virtue of its incompatability with the officially reported election results.
I cannot find this data. Could someone please point me towards it?