berkeley-hout paper mit summary und data: Working Paper: The Effect of Electronic Voting Machines on Change in Support for Bush in the 2004 Florida Elections by Michael Hout, Laura Mangels, Jennifer Carlson, and Rachel Best

dazu auch: Scholars on the votes, Ohio undervotes

und salon.com: It’s not proof of voter fraud -- at least not yet -- but it seems that somebody has some explaining to do about the election results from Florida. In a report released this morning, researchers at the University of California, Berkeley, say that George W. Bush received 130,000 more votes in Florida in 2004 than he should have received, and that the only real explanation has something to do with electronic voting machines.

Hout said the researchers applied their same tests to electronic voting in Ohio and discovered no such disparities. And even if the Berkeley researchers are right about Florida, their numbers don't change the overall result of the election there. As things stand now, Bush won Florida by about 311,000 votes. If the 130,000 "extra" votes the Berkeley researchers have found were "ghost votes" – that is, votes that were never cast but simply added to Bush’s total – then Bush's margin would drop to about 181,000 votes. But if the 130,000 votes were Kerry votes that somehow got switched to Bush votes, then Bush’s margin in Florida would drop to 51,000.