Article

E-voting hits all-time high in '04, but skeptics voice concern

The use of electronic voting systems was more widespread than ever on Election Day '04. And leaders in the industry expect the use of touchscreen voting machines to increase over the next four years. But some computer technologists and scientists have voiced concerns about some machines' lack of paper ballots or paper trails.

November 21, 2004 by Tracy Kitten — Editor, AMC

The use of electronic voting machines on Election Day 2004 was unprecedented. Thirty-one states and the District of Columbia used touchscreen e-voting machines or kiosks. Twenty-nine percent of the nation's voters used touchscreens this year, the greatest percentage ever, according to Verifiedvoting.org, which is managed by the Verified Voting Foundation and Computer Professionals for Social Responsibility.

Given those high numbers, it's not surprising that some concern over the use of touchscreen e-voting has surfaced.

Skeptics claim, with touchscreen e-voting, that there's no way to guarantee votes are being counted accurately or at all. Supporters of the touchscreen method claim their adversaries are too caught up in conspiracy theories and unlikely scenarios to see the practicality of touchscreen voting.

The fact is, however, this year's U.S. Presidential Election went relatively smoothly, according to elections offices throughout the nation.

In Florida, where a lot of media coverage was focused, Election Day came to a close with relatively few noted problems, according to Jenny Nash, press secretary for the Florida Department of State.

"Florida started using the touchscreen voting in 2002," Nash said. "Literally, in Florida, there is an election somewhere every week Â… and we've had hundreds of successful election systems using this method. We've never had reports of equipment malfunctions (or votes being lost)."

Florida has been using optical-scanning machines, which scan and tally paper ballots, since 2000.

Nash said 15 of Florida's 67 counties used touchscreens this year, and approximately half of the state's voters reside in those 15 counties. The majority of the state, she added, uses touchscreens from Sequoia Voting Systems and the Electronic Systems & Software iVontronic machine. One of Florida's counties, Duval County, uses a touchscreen from Diebold Election Systems. And most counties in Florida use optical scanners from all three companies, Sequoia and ES&amp's and Diebold, Nash added.

Florida has been using the Diebold scanning machines since 2002. And 30 counties currently use the Diebold optical scanning machines, she said.

"The use of touchscreens is very important, because, for the first time ever, visually impaired and illiterate people, and even people who are paralyzed, can use this system to vote without assistance," Nash said.

By January 2006, under the federal Help America Vote Act of 2002, every precinct in the nation will be required to have at least one touchscreen voting machine (or any machine with the same capabilities) that allows disabled individuals to cast independent ballots.

Nash added that under-voting - when a voter neglects to choose a candidate for a particular race - is at an all-time low, about 1 percent. Over-voting - when a voter selects more than one candidate for a particular office -she said, is a non-issue, since touchscreens will not accepts over-votes and optical-scanning machines alert poll workers and voters of errors before voters leave the precinct. "The optical scan machines alert the voter (when an over-vote is cast) and allow them to correct their ballot," Nash said. "But if the voter insists on casting the ballot as is, they may do so."

The advantages are obvious, Nash said; but, as others argue, so are the disadvantages.

Will Doherty, a computer technologist and executive director of the Verified Voting Foundation, said states need to take precautions. The Verified Voting Foundation, which was initiated more than a year ago, is focusing its efforts on eliminating tampering and fraud opportunities associated with e-voting.

Citing an incident in Cateret County, N.C., where more than 4,000 early votes or "One Stop" ballots were lost on a UniLect Corp. Patriot Voting System machine, Doherty argued it's obvious that voting machines aren't infallible.

According to UniLect president Jack Gerbel, UniLect admits it erred in telling Cateret County poll workers that the machine could retain more votes than was realistic. Poll workers were told that the machine's black box, which stores the votes, could hold about 10,000 votes. In reality, the black box retained only 3,005.

"It was very unfortunate, and we feel extremely sad that it happened," Gerbel said. "It (the machine) had more votes than the machine could hold. Â… But they're one of our oldest customers and their machine just hadn't been updated."

Gerbel said all of UniLect's other Patriots have been updated, however.

But Doherty said mistakes, like the oversight of updating the Cateret County machine, can happen anywhere, and elections officials need to have a back-up system in place. In fact, according to Verifiedvoting.org, 1,691 e-voting machine problems were reported on Election Day.

"The main concern we have is that e-voting machines are being used nationwide without paper ballots," Doherty said. "What we want to see is a voter-verified paper ballot. We're not talking about a receipt. We're not talking about something you would take home with you. We're talking about something that can be used for audits and recounts. If there's a problem, you could go back and check."

Nash said she doesn't see why any paper during a vote is needed, however, since all of the machines used in Florida are "rigorously" tested and certified by the Florida Department of State; and all of the machines are capable of printing an audit trail at the end of the day.

"Florida not only has one of the most rigorous certification processes in the nation, but prior to each and every election, a pre-election logic and accuracy test is conducted on every voting machine that will be used in the election," Nash said. "After testing, the machines are then sealed and that seal is not broken until Election Day."

Nash added that "Florida voters should have complete confidence in these voting systems, due to the rigorous certification process and pre-election testing."

Greg Swistak, executive director of the Kiosks.org Association, agrees.

"The whole notion that you'd have a paper record creates more controversy," Swistak said. "Everything is going away from that (to less paper), and as far as I'm concerned, the less paper you have, the better. The elections were very positive in favor of the machines. People like using them. And as those machines are used more and more, they will be received favorably."

Swistak added that the concerns Doherty raised are difficult to realistically conceive.

"He's saying that you'd have to manipulate the machines at the time they were built in the factory to change the outcome for a certain candidate at a certain time in the future," Swistak said. "This would require you to anticipate who the candidates will be and, also, whom votes would be shifted to. The vote-collecting code itself would have to be designed to do this because the candidates' names and positions on the screen are random.Â… That's like saying when you pull the lever (on a mechanical voting machine), someone figured out a way to have the vote go to someone else."

But Doherty said it can happen. To illustrate the concern, Doherty pointed to "Analysis of an Electronic Voting System," a report that includes study findings from Johns Hopkins University. The report outlines ways the security, accuracy and functionality of an electronic voting system can be compromised. Doherty went on to say the catalyst for the report - the discovery of software source code for some Diebold Inc. systems on the Web - is a definitive reason for e-voting concerns.

Diebold has refuted many of the report's findings, according to David Bear, spokesperson for Diebold Election Systems. "That report was based on some outdated and incomplete code that made some incorrect assumptions," Bear said. "One thing they said was that the system could be hacked, but they didn't take into consideration that these are stand alone systems. And they didn't test the theories in an election environment -they didn't take into consideration the checks and balances."

Doherty, on the other hand, said some states ran independent tests after the report came out and found similar bugs. "The report caused the state of Maryland to conduct two more independent investigations of Diebold machines," Doherty said. "They (Diebold) claimed they had fixed everything, but they hadn't (according to Maryland's findings)."

Bear said Maryland felt confident enough after running two independent tests to purchase Diebold's AccuVote-TS for its state's elections. The tests, however, were run on the AccuVote-TSX.

"Diebold's machines have probably been independently studied more than any other machines in the market," Bear said. "The election officials are very pleased with the performance system Â… because they're much more adaptable to ADA (Americans with Disabilities Act) and everyone can access the system."

Bear added that Diebold has optical-scanning and touchscreen machines in about 26 states, 14 of which primarily use AccuVote-TS machines.

Meaghan McCormick, spokesperson for ES&S, said she doesn't see skepticism of e-voting machines in a negative light. In fact, she said her company appreciates the scrutiny.

"We want people to think," McCormick said. "That's not a bad thing, to have people paying attention."

McCormick said the ES&amp's iVotronic, which is used in 20 states, has three separate memory chips that are used as a way to "check" and "balance" vote tallies. Using the three chips ensures that different information is saved in different places, she said, and that printouts can be produced for recounts.

"We've never had any accusation or anything reported about security breaches," she said. "This is a proprietary system. It's not connected to a network and there are so many checks and balances in place. Â… And we absolutely maintain that all of our systems are secure."

Doherty maintains, however, that counties need to take extra precautions, regardless of the fact that most voting machines are free-standing kiosks.

"Software in the machines could be changed anytime, anywhere during the day," he said. "So the only way to double check votes is to have a paper ballot box. Most (computer scientists and technologists) agree that we need to have paper ballots. Â… Even though it's a standalone unit Â… an inside attack (from someone within the company who designs the software) could cause certain things to happen (through the way the code is written).If the party affiliation is indicated on the ballot, then someone could simply attack by party affiliation."

Doherty said the code could be manipulated by a disgruntled for a variety of reasons, not just because he or she wants a certain candidate to win.

"I just really don't see how anyone could argue that you don't need paper ballots," Doherty said. "Look at an ATM machine. It prints out a receipt for a person and then a receipt for the bank. Why do we do that? Why do we have a $20 ATM transaction handled more carefully than our right to vote, in our Democratic society?"

What voters want

The jury is still out on what voters want. According to David Bear, spokesperson for Diebold Election Systems, Diebold receives approval percentages in the mid- to high- 90s for its e-voting machines. Bear and Jenny Nash, of the Florida Department of State, said the flexibility of touchscreens, where language, accessibility and various ballot options (for counties that have numerous ballot styles) are concerned, makes them attractive to elections officials.

In a nutshell, touchscreens are more convenient for voters. Nash added that "people in Florida are familiar and comfortable with them."

However, a survey conducted by Ball State University's Center for Media Design, which canvassed 421 voters in Muncie, Ind., just days before the election, showed different results. According to the Ball State study, 33 percent of the voters surveyed preferred paper ballots over 16.4 percent who preferred voting on an "electronic kiosk." In that survey, voters did, however, show an interest in some form of e-voting - 28 percent said they would prefer voting on the Internet.

And an Election Day poll conducted by Lombardo Consulting Group LLC found among the 362 voters in Las Vegas who were surveyed about touchscreen voting, 81 percent preferred being given private take-home, "ATM-style" receipts to verify their votes. When asked to make a choice between leaving a voter-verified paper ballot at the polling place over taking a voter-verified receipt home, 60 percent said they preferred taking a receipt home while 36 percent said they preferred the idea of leaving a paper ballot with the election officials.

Nevada was the only state to use statewide voter-verified paper-ballot printers that were attached to the e-voting machines.