Take another look at those Dixiecrats

Salon may or may not publish my response to Farhad Manjoo's skeptical new piece on voter fraud. In that response, I decried the continued attention paid to Kathy Dopp's chart showing odd results in Florida's "Dixiecrat" counties. I argued that her figures divert us from the more important issue of the exceptionally vulnerable central tabulator.

However, on Air America today, Randi Rhodes -- who, unlike Manjoo or myself, lives in Florida -- discounted the oft-heard theory that "blue dog" Democrats in those small counties are so numerous as to justify the massive party-switching in favor of Bush. Thom Hartmann takes that argument further here.

At this point, I'm not quite sure wher to stand. Perhaps we should hear out the other side to the story:

One possible explanation for this is the "Dixiecrat" theory, that in Florida white voters (particularly the rural ones) have been registered as Democrats for years, but voting Republican since Reagan. Looking at the 2000 statistics, also available on Dopp's site, there are similar anomalies, although the trends are not as strong as in 2004. But some suggest the 2000 election may have been questionable in Florida, too.

One of the people involved in Dopp's analysis noted that it may be possible to determine the validity of the "rural Democrat" theory by comparing Florida's white rural counties to those of Pennsylvania, another swing state but one that went for Kerry, as the exit polls there predicted. Interestingly, the Pennsylvania analysis, available at http://ustogether.org/election04/PA_vote_patt.htm, doesn't show the same kind of swings as does Florida, lending credence to the possibility of problems in Florida.

Even more significantly, Dopp had first run the analysis while filtering out smaller (rural) counties, and still found that the only variable that accounted for a swing toward Republican voting was the use of optical-scan machines, whereas counties with touch-screen machines generally didn't swing - regardless of size.

Others offer similar insights, based on other data. A professor at the University of Massachusetts, Amherst, noted that in Florida the vote to raise the minimum wage was approved by 72%, although Kerry got 48%. "The correlation between voting for the minimum wage increase and voting for Kerry isn't likely to be perfect," he noted, "but one would normally expect that the gap - of 1.5 million votes - to be far smaller than it was."

An even more surprising figure: Bush earned 20,000 more votes statewide than the sum total of the state's registered Republicans.