Electronic Voting A Mess In March 2004 California Primary

The following articles are reproduced under the Fair Use exception of 17 USC § 107 for noncommercial, nonprofit, and educational use.


| EJF Home | Where To Find Help | Join the EJF | Comments? | Get EJF newsletter |


| Vote Fraud and Election Issues Book | Table of Contents | Site Map | Index |


| Chapter 10 — Voting Problems In The 2004 Elections |

| Next — North Carolina Takes The Lead In November 2004 Election Problems |



Electronic voting bug plagued San Diego County election

Electronic voting component wasn't fully tested

Lost electronic votes could flip Napa County race


Electronic voting bug plagued San Diego County election by Michelle Morgante

Independent Media TV

March 12, 2004, San Diego (AP) — A computer battery problem affected about 40 percent of polling stations in San Diego County, delaying and frustrating voters who lined up to cast electronic ballots in last week's primary election, according to a county report issued Wednesday.

In the largest rollout of an e-voting system by any local jurisdiction in the nation, San Diego County officials believe that the problems prevented a number of people from casting votes. "There is no method to accurately measure how many voters were unable to vote," the report said.

On March 2, San Diego County was quickly overwhelmed with calls for help when poll workers turned on electronic devices that encode the magnetic-striped cards used to access touch-screen machines.

Poll workers were trained to expect their computer screens to show a page from the voting-system software. Instead, 40 percent of the 1,611 devices initially displayed a screen from the Windows operating system, according to the report by the county's Chief Administrative Office.

Though only four computer clicks were needed to advance to the expected screen, many poll workers had no idea how to do so. The "widespread problem" with the encoding devices "was considered to be 'low probability,'" so poll workers were not given instructions on how to reach the proper screen, or alternative methods for encoding the access cards, according to the report.

"At a few locations, voters actually assisted poll workers in maneuvering through the startup process to reach the login screen," the report stated.

Diebold Election Systems of McKinney, Texas, is trying to determine the root of the problem. The county's report blamed an unexpected discharge from an internal battery that caused the computers to reset themselves and display the Windows screen.

"We just don't know yet why there would have been a low battery or power-source issue," Diebold spokesman David Bear said. "We are certainly looking at it."

The county said Diebold is expected to deliver its own report on the problem in about two weeks.

In Alameda County, which used the same voting machines made by Diebold, about 200 of the county's 1,096 polling stations saw a variety of problems with the encoding devices. Alameda County gave affected voters paper ballots or asked them to return later. Officials expect to conclude a report on the problems by the end of the month, said Elaine Ginnold, the assistant registrar of voters.

The Los Angeles Times found that about 7,000 Orange County voters were given the wrong ballots by poll workers struggling to figure out their new e-voting system, which was made by a different company, Hart InterCivic.

In San Diego County, some 6,800 poll workers were recruited for the March 2 election. County spokeswoman Linda Miller said poll workers with computer skills will be needed for future elections.

The county Registrar of Voters had 11 troubleshooter hotlines set up, as well as 50 other phone lines available for poll workers to reach help. But the lines couldn't handle the rush of calls when polls attempted to open simultaneously at 7 AM.

The report said 64 percent of polling stations managed to open on time. By 8 AM, 88 percent were open, and by 9 AM, all but 31 polling stations were open. The last station opened by 11:05 AM.

San Diego County, like others across the state and country, was forced to change its election system because of the 2000 presidential fiasco in Florida, which led California and federal officials to ban the old punch-card ballots that were plagued by hanging chads and other problems.


Electronic voting component wasn't fully tested by Helen Gao


© 2004 by Helen Gao, San Diego Union-Tribune

March 13, 2004 — State election officials knew in the days leading up to the March 2 election that a key component of San Diego County's electronic voting system had not undergone the full testing set forth by federal standards, according to internal government correspondence.

That component, a laptop-like device used to activate voter cards to call up ballots on touch screens, failed on Election Day because of a battery problem.

The glitch caused 36 percent of the precincts, or 573 of 1,611, to open late. An undetermined number of voters were affected.

County officials were informed that the device only had conditional certification for one-time use on March 2, election day, because of outstanding testing issues, according to a February 20 letter the county provided yesterday.

The letter from the state to Diebold Election Systems, the manufacturer of the device, said that after the election, the equipment needs to be resubmitted for certification.

The state gave conditional certification to the device after its own technical consultant recommended approval in a Feb. 23 letter, after what he said was a "limited" review of "limited" testing by an independent laboratory.

Board of Supervisors Chairwoman Dianne Jacob said the county relied on the state certification and did not know about the state consultant's letter.

"If we were in the secretary of state's position and had that information, I am not sure if this county would have made the same decision" to certify the device for use, she said.

"The secretary of state did the certification. Based on that, we are forced to trust his judgment."

The state did not do more testing because of the "urgent March deadline," said consultant Steven V. Freeman in his letter to the state's director of voting systems.

The testing focused only on whether the device could encode voter cards properly according to precinct and political party.

Freeman did not raise concerns about the battery in his letter, but he did point out that the device had not been fully tested according to the Federal Election Commission's Federal Voting Systems Standards.

The standards, which California adopted, contain technical specifications to ensure electronic voting systems are "accurate, reliable and secure."

Full testing would have required checking the performance of the equipment under normal and abnormal conditions. In addition, it would have required a series of evaluations of the software and hardware to ensure the equipment holds up during storage, operation and transportation.

Doug Stone, spokesman for Secretary of State Kevin Shelley, blamed Diebold for submitting the equipment late for testing. He said it wasn't until January that Diebold made the equipment available.

"From our standpoint, Diebold dragged its feet in this process and frankly, it was quite frustrating," Stone said.

He said the state certified the device because testing by independent laboratory CIBER Inc. showed that the equipment functioned.

Diebold's spokesman, David Bear, defended the device. "It did meet CIBER testing and met approval from the Secretary of State's Office," he said.

When sought later for comments on Diebold's delay in submitting the device, Bear did not return calls.

Kim Alexander, president of the California Voter Foundation, which has repeatedly raised concerns about the security of electronic voting, said state and federal officials need to do a better job of oversight.

"This incident...shows how the regulatory oversight of voting systems is a house of cards and it's easily toppled," Alexander said.

The state, jittery about electronic voting in general, on February 11 ordered San Diego and other counties to provide paper ballots as an option March 2.

By then, county officials said it was too late to print the different types of ballots needed and distribute them to all the precincts. The county had about 33,000 ballot types, including language translations.

In the end, the county provided paper ballots only at the registrar's office in Kearny Mesa.

The March 2 election, being a modified open primary, was complicated. Up until October, different parties could decide whether they would allow nonpartisans to cross over and vote in their races.

"It was not entirely Diebold's fault that (the device) came so late. Before any testing could be done, Diebold had to know what those party rules were" to program the equipment, said Elaine Ginnold, Alameda County's assistant registrar.

Alameda, which uses Diebold, also had equipment problems but nearly all its polls opened on time because they had provisional paper ballots. San Diego County opted for electronic provisionals.

Ginnold said state lawmakers share some responsibility for setting rules that make elections increasingly complicated.

"There is just not enough time to program something and do all the testing that is required," she said.


Lost electronic votes could flip Napa County race by Kim Zetter


© 2004 by Kim Zetter, Wired News

March 15, 2004 — Napa County in Northern California said on Friday that electronic voting machines used in the March presidential primary failed to record votes on some of its paper ballots, which will force the county to re-scan over 11,000 ballots and possibly change the outcome of some close local races.

The glitch is the latest in a string of problems with the new generation of electronic voting machines being rolled out across the United States. Critics of the machines say they are inaccurate or susceptible to tampering, and can't be trusted in this year's presidential elections.

The problem occurred with optical scan machines manufactured by Sequoia Voting Systems, which failed to record voters' marks off of paper ballots. The county used the company's Optech system for processing paper absentee ballots.

Napa Registrar of Voters John Tuteur said they discovered the problem on Thursday while conducting a manual recount of 1 percent of precincts, to verify accuracy, a statewide practice. Tuteur said after counting a sample of 60 paper ballots from one precinct, officials discovered that the number of votes did not match the number of votes the machine recorded for that precinct. After re-scanning 10 of the ballots, they discovered that the machine wasn't recording certain votes.

Sequoia spokesman Alfie Charles said the problem wasn't with his company's machines. "It was a procedural error on the part of the people who were setting up the equipment," he said.

Specifically, the machine was calibrated to detect carbon-based ink, but not dye-based ink commonly used in gel pens, Charles said. Prior to the election, a Sequoia technician ran test ballots through the machine to calibrate its reading sensitivity, but failed to test for gel ink.

"The problem was isolated to the one machine in Napa and was detected and properly calibrated within hours of identifying it," Charles said. "It's important to note that the check and balances in place worked," referring to the required manual recount.

Kim Alexander, president and founder of the California Voter Foundation, a nonprofit voter education organization, said the county was lucky that the problem occurred on a system with a paper trail.

"If the problem had occurred with their electronic ballots or with the tabulation software (that sits on the county server) they would have been hard pressed to reconstruct their election," she said. "Or they might not have ever known there was a problem at all. If they were doing the manual count on the electronic ballots there would be no record to look at to determine what the accurate vote count should be."

She added California is "one of a few if not the only state" that requires a hand count.

"The reason we have the manual-count verification is precisely because technology is not always reliable. There have been many instances like this where the manual count has been instrumental in flagging a vote counting problem," she said.

Tuteur said that as soon as the Sequoia technician recalibrates the machine, the county would re-scan all the paper ballots.

At least one close race could be overturned. Incumbent county supervisor Mike Rippey narrowly lost his re-election bid by only 50 votes.

"At this point in time we have no confidence in the results coming out of these machines," said Rippey's campaign spokeswoman Linda Scott. "What concerns us the most is that the count is so close on the absentee ballots that it could sway the election results."

The primary was the third time the county had used the Sequoia machine, Tuteur said.

"We don't know if this problem has occurred before but we' re not aware of any other problems," he said.


To read Wired News' complete coverage of e-voting, visit the Machine Politics section.



| EJF Home | Where To Find Help | Join the EJF | Comments? | Get EJF newsletter |


| Vote Fraud and Election Issues Book | Table of Contents | Site Map | Index |


| Chapter 10 — Voting Problems In The 2004 Elections |

| Next — North Carolina Takes The Lead In November 2004 Election Problems |


Last updated 6/14/09