Progress Pond

Unverifiable: The South Carolina Primary

Cross-posted at Open Left and Daily Kos

The South Carolina primary could well be a do-or-die contest for at least one of the three top candidates in both major parties. Depending on how the first three states shake up, Huckabee, Romney, or Giuliani will be faced with on-the-ropes status in news coverage by the January 19 primary. Failure to win there could subject one of them to a full-blown candidacy death watch. The same goes for Clinton, Edwards, and Obama and the January 26 Democratic primary in South Carolina. Not to mention the rest of the field.

But the South Carolina primaries will not verifiable. Both will be conducted on a statewide paperless electronic voting system which has become notorious among activists and computer scientists for its reliability and security issues.

Meet the ES&S iVotronic.
The Republican Party of Wharton County, Texas decided just last week that the the iVotronic is not reliable enough to use in its March 2008 primary election. Wharton County Republican Party chairwoman Debra Medina announced that the party’s 22 precinct chairs had agreed to use voter-marked paper ballots counted by optical scanners.

Apart from South Carolina, counties in 10 other states will also use the iVotronic in next years’ primaries.

The Wharton County Republicans’ decision followed an incident in the November 7 election in which a local businessman tried to vote on one proposed constitutional amendment, and saw his previous vote for another proposed amendment change before his eyes.

The iVotronic is the machine of Sarasota 2006 fame, producing implausible undervotes in Florida’s 13th Congressional District, as well as high undervotes in other races in six Florida counties that used the machines. It is the same machine whose firmware version 8.0.1.2 was described by Princeton University computer scientist Edward Felten as “terribly insecure” and in need of serious improvements before it used in another election.

It is the same machine that flipped votes in Texas, Indiana, Pennsylvania, Florida, and yes, South Carolina, in 2006:

Douglas Jones, a computer scientist at the University of Iowa, says he’s heard similar stories from voters in several states, including one computer scientist in South Carolina who said that his attempts to vote for one candidate on the iVotronic were repeatedly changed to an opposing candidate by the time he got to the voter verification screen.”

A note about the above: vote-flipping on the selection screen can be explained by the screen’s sensitivities falling out of sync with the candidate display positions. That is the so-called “calibration issue.” Deliberate miscalibration can be a form of an attack, as the above-quoted University of Iowa computer scientist Douglas Jones observed in 2005.

Vote-flipping on the review screen cannot explained by calibration problems. A calibration slippage would reveal itself when the voter chose a candidate on the selection screen, not at the last stage of the voting process. Most alarming: a study done by Rice University graduate students found that 60% of subjects failed to notice vote alterations on the review screen.

The iVotronic is the machine discussed in Dan Rather’s special, “The Trouble With Touch Screens”  in August of this year. Manufactured in sweatshops (a fact which, unlawfully, was not disclosed to the Election Assistance Commission), and with abysmal quality control.

In all, 13 states appear set to run Presidential primaries that use paperless machines either exclusively or extensively.  These include Virginia, Florida, and Maryland, which have passed legislation to end paperless voting but are in various stages of transition.

In addition to the iVotronics, paperless states will also use equipment manufactured by Diebold and Sequoia, machines found to have severe security vulnerabilities by an independent review team commissioned by California Secretary of State Debra Bowen. Bowen has judged the Diebold and Sequoia systems too insecure for use as a primary voting system in her state’s February 5 primary. A Diebold system was hacked in 2006 by Princeton computer scientists, who were able to write a vote-switching program that could pass rigorous pre-election testing and spread from machine to machine like a virus. The AVS WinVote was recently disallowed for use in Pennsylvania after almost 2000 source code anomalies were uncovered during federal certification testing. The company’s application for certification was terminated by the Election Assistance Commission last week, but voters in Virginia and Mississippi will still use them to cast votes next year.

But no matter what system is used, paperless electronic voting is inherently dangerous to democracy, as thousands of computer scientists and other technologists have concluded

Election officials and party leaders in the paperless states have an unenviable job.  They were promised that the equipment they bought in the last four years was reliable and secure, and are gradually realizing that they were told wrong. But now they have primary elections to run.  Should they trust to hope that the votes are recorded accurately, or should they put themselves through what would no doubt be a difficult process of implementing a paper ballot contingency?

Time is short, but there is no fair alternative: no voter in the Presidential primaries should be required to trust his or her vote to a paperless machine.

What to do? Contact the South Carolina Electoral Commission, and the state Democratic and Republican parties, for starters. Time is short, to say the least.  But citizens who care about verifiable elections have to make their voice heard.

The Wharton County Republicans’ Medina said it best:

“I don’t want to be on the front page of any newspaper having to say our vote (was) unreliable,” she said. “We work very hard to get voters to the polls, and if we can’t rely on the vote to be the intent of the people, what are we doing?”

0 0 votes
Article Rating
Exit mobile version