Florida voting audit raises more questions

The audit of voting machines begun yesterday in Sarasota County, Florida was supposed to help clarify when happened to the 18,000 ballots which "undervotes" that didn't register a choice for Congress. But puzzling glitches and criticisms of the audit itself means the battle over the U.S. House seat, in which Republican candidate Vern Buchanan was declared the winner by a 369 vote margin, will likely continue.

As the Bradenton Herald reports:
The results of a simulated election held on spare machines didn't exactly match what state elections officials had scripted. Of four machines, the results of which were compared to actual election-day results, three had one to three fewer votes for Democrat Christine Jennings in the 13th Congressional District race.
The responses of the parties involved were predictable. Election officials, who happen to be Republican, called the discrepancies "unsurprising" and the result of "human error." Democrat Christine Jennings' campaign called them "intriguing." Bachanan's rep said it proved the Jennings' case "doesn't hold water."

Voter advocates have contended that the audit is flawed in its very design, and therefore is unlikely to be a useful tool in revealing what happened to the missing votes:

Tuesday's test was designed by state elections officials and an expert retained by Jennings to "duplicate election day as much as possible," secretary of state spokeswoman Nash said.

But critics said it was anything but, citing numerous differences.

The touch-screen machines were hung vertically instead of lying flat as they were on Election Day. Most who cast ballots in the Nov. 7 election did so once in private while standing up, not repeatedly in plain view of onlookers and video cameras while sitting down.

And the test teams, whose members appeared to range in age from the mid-20s to early 50s, also didn't accurately represent real voters, critics said.

"They're professional testers," Coffey said of the state employees.

Critics also said the 10 machines to be used in the mock elections are too few to yield any meaningful results. The county used 1,498 machines in the real election.