Palm Beach eVoting Logs Call Vote Into Question
Sunday February 26, 2006
Analysis of 2004 electronic voting logs from Palm Beach County (FL) reveals 100,000 errors, including 1,475 instances of recalibration during polling hours and votes cast on days other than election day. Recalibration occurs when a machine is malfunctioning. Sequoia Systems, which makes the equipment, defended its performance, as did Arthur Anderson, county elections supervisor. However, because of the manner in which evoting data are saved, it is impossible to determine whether or not there was foul play, intentional or accidental.
About 10 percent (40 machines) show strange dates, according to Bev Harris of Black Box Voting. Election officials say that workers merely set the date incorrectly. However, "Harris said the logs show that 40 machines had the correct date and time during routine tests a few weeks before the Nov. 2, 2004, election. But sometime after the tests, Harris said, other dates and times showed up. Most of the dates were in late October 2004, but one machine showed Nov. 2, 2010, Harris said."
The logs didn't come "free" -- they cost Harris more than $7,000. What activist has the money to pay for logs from every eVoting precinct in the country? Why are these data not public information?
Could it be because the companies claim the information is company secret? That seems to be the case in Alaska, where a public records request for Diebold records has been denied by the Division of Elections as a "security risk." Earlier, the agency had agreed to release the data, but dragged its feet. The Democratic party is seeking the records from the 2004 election; there are reports of 200 percent turnout in some precincts.
All the more reason to vote with paper ballots that can be recounted with human eyes. Recounting flawed zeros-and-ones is an exercise in futility, if the goal is to "prove" the vote count matches voter intent. Florida does not require paper trails on touch-screen voting machines.
Stay on top of these issues at Vote Trust USA.
About 10 percent (40 machines) show strange dates, according to Bev Harris of Black Box Voting. Election officials say that workers merely set the date incorrectly. However, "Harris said the logs show that 40 machines had the correct date and time during routine tests a few weeks before the Nov. 2, 2004, election. But sometime after the tests, Harris said, other dates and times showed up. Most of the dates were in late October 2004, but one machine showed Nov. 2, 2010, Harris said."
The logs didn't come "free" -- they cost Harris more than $7,000. What activist has the money to pay for logs from every eVoting precinct in the country? Why are these data not public information?
Could it be because the companies claim the information is company secret? That seems to be the case in Alaska, where a public records request for Diebold records has been denied by the Division of Elections as a "security risk." Earlier, the agency had agreed to release the data, but dragged its feet. The Democratic party is seeking the records from the 2004 election; there are reports of 200 percent turnout in some precincts.
All the more reason to vote with paper ballots that can be recounted with human eyes. Recounting flawed zeros-and-ones is an exercise in futility, if the goal is to "prove" the vote count matches voter intent. Florida does not require paper trails on touch-screen voting machines.
Stay on top of these issues at Vote Trust USA.
Technorati Profile
Technorati tags:
eVoting,
Politics
gada.be tags:
eVoting,
Politics