11.26.2006

Paul Krugman Agrees: "When Votes Disappear"

Paul Krugman goes into far better detail about the phenom I've posted here should have us ALL concerned: just because a few Dems won on November 7th does NOT mean the voting process worked. It just means it "somewhat worked".

Whenever any vote goes missing, gets flipped to the opponent, or otherwise fails to get counted properly, we all lose.

From Krugman:

You know what really had me terrified on Nov. 7? The all-too-real possibility of a highly suspect result. What would we have done if the Republicans had held on to the House by a narrow margin, but circumstantial evidence strongly suggested that a combination of vote suppression and defective — or rigged — electronic voting machines made the difference?

Fortunately, it wasn’t a close election. But the fact that our electoral system worked well enough to register an overwhelming Democratic landslide doesn’t mean that things are O.K. There were many problems with voting in this election — and in at least one Congressional race, the evidence strongly suggests that paperless voting machines failed to count thousands of votes, and that the disappearance of these votes delivered the race to the wrong candidate.

Here’s the background: Florida’s 13th Congressional District is currently represented by Katherine Harris, who as Florida’s secretary of state during the 2000 recount famously acted as a partisan Republican rather than a fair referee. This year Ms. Harris didn’t run for re-election, making an unsuccessful bid for the Senate instead. But according to the official vote count, the Republicans held on to her seat, with Vern Buchanan, the G.O.P. candidate, narrowly defeating Christine Jennings, the Democrat.

The problem is that the official vote count isn’t credible. In much of the 13th District, the voting pattern looks normal. But in Sarasota County, which used touch-screen voting machines made by Election Systems and Software, almost 18,000 voters — nearly 15 percent of those who cast ballots using the machines — supposedly failed to vote for either candidate in the hotly contested Congressional race. That compares with undervote rates ranging from 2.2 to 5.3 percent in neighboring counties.

Reporting by The Herald-Tribune of Sarasota, which interviewed hundreds of voters who called the paper to report problems at the polls, strongly suggests that the huge apparent undervote was caused by bugs in the ES&S software.

About a third of those interviewed by the paper reported that they couldn’t even find the Congressional race on the screen. This could conceivably have been the result of bad ballot design, but many of them insisted that they looked hard for the race. Moreover, more than 60 percent of those interviewed by The Herald-Tribune reported that they did cast a vote in the Congressional race — but that this vote didn’t show up on the ballot summary page they were shown at the end of the voting process.