Out of Touch
You press the screen. The machine tells you that your vote has been counted. But how can you be sure?
BY WYATT OLSON

Emil Danciu says he smelled victory a year ago. He's slouched in his chair and looks a bit out of place seated in front of his daughter's office desk. He's faced with a phalanx of framed baby photos. Charlotte Danciu, a Boca Raton attorney who specializes in adoption and juvenile law, has decorated the corner office with scores of framed snapshots of tots and adoring parents.

Her father, a calm man of 72, is dressed casually in shorts, polo shirt, and sneakers.
[...]
Early last year, Emil Danciu was one of four candidates vying to win two seats on the Boca Raton City Council. A long-time civic activist, he had spearheaded the drive to preserve public land in Boca Raton in the '80s. He served as mayor from 1985 until 1993, when he lost by 99 votes. In Boca's polarized political camps of development versus slow growth, Danciu represented the latter.

Danciu lost that race, an election mired by glitches in the county's new touch-screen voting computers. Later that fall, during the primary, similar machines by a different manufacturer exhibited some of the same problems in Broward County. Now, a growing number of computer scientists are sounding the alarm over the nation's growing reliance on computerized voting machines that provide no paper trail.

Some experts contend that the Danciu election underscores the problems in dealing with paperless recounts and the dubious reliability and security of this new technology. At the same time, exit polling by media outlets -- which is one of the few methods available to detect ballot-box anomalies -- is waning. The nation's switchover to touch-screen computers will likely speed up as billions of dollars authorized by Congress become available for the new technology.

"I quite definitely thought I'd win," Emil Danciu recalls of the election held on March 12, 2002. "We had a good group put together, and we had all the supposedly good things you're supposed to have in a campaign. We had a decent amount of money."

His daughter interjects: "I was the unofficial campaign manager, I guess you could say. All of the polling that was done indicated that he'd win because of this antidevelopment sentiment that was being expressed in Boca. But the real telling factor was that all of the big developers were calling him the weekend before the election trying to get on his good side, offering him money, offering him signs."

The Dancius learned that an opposing camp had conducted a poll about two weeks prior to the election and found Danciu about 17 points ahead. "Like everybody else does, we had a mole in the other operation," Charlotte explains. "This poll was taken by the Chamber of Commerce people, the people with all the money."

Even the exit polling they conducted on Election Day was encouraging. "They were overwhelmingly indicating that he would be the victor," she says.

The Dancius gathered at their headquarters on Federal Highway on election night with friends and campaign workers. It was the first election to employ the new touch screens made by Sequoia, which had been touted as offering quick returns. "We kept waiting and waiting, but Boca couldn't get any results out," Charlotte says. Finally, she and a few others drove to the office of Theresa LePore, the supervisor of elections for Palm Beach County. "There was an army of reporters and an army of people from Sequoia," Charlotte says. "At that point, they said they couldn't tabulate the votes because they'd lost 15 cartridges. They were just missing, and the system was built so it wouldn't give a final tally until you turned in all the cartridges. Then they claimed a poll worker had taken them home, and then they found them."

The Dancius were stunned when the totals did come in: Danciu had received 2,863, which was a few thousand votes short of the two winners, Susan Haynie at 6,044 and Bill Hager at 5,446. Even more perplexing was the fact that Danciu had lost in his own home voting precinct.

For about a decade, Emil Danciu says, the slow-growth candidates had been "on the short end by a couple hundred votes, at most. All of a sudden, here's a 2,000-vote spread between us and them. That was really strange to us."

Charlotte continues: "What really alarmed us was the next day when we started getting phone calls from voters who had gone into the voting places -- people we didn't even know -- and pushed Emil Danciu's name only to end up with a check mark by Susan Haynie's name. They repeatedly tried to vote for him, but another name, particularly Haynie's, came up. They couldn't get their vote registered. They were telling wild stories about poll workers unplugging and kicking the machines. They didn't know whether their votes ever counted. Some were told to vote again."

Haynie and Hager were certified as winners by LePore's office. Charlotte Danciu, then joined by lead attorney Rob Ross, filed suit contesting the results on March 25, 2002, in Palm Beach County Circuit Court.

Their case was bolstered, they believed, when Councilman Al Paglia lost his seat by four votes to Lizbeth Benacquisto during a runoff contest held March 26 in Wellington, a town of 42,000 in central Palm Beach County. Although Paglia and Benacquisto were the sole candidates on the ballot, 78 so-called undervotes were registered, meaning 78 voters used the machine but did not cast a ballot. That struck Paglia as odd because he'd garnered 45 percent of the votes during the primary run against three challengers. And then, he too began hearing stories from voters that the Sequoia touch screens had acted erratically. On Paglia's behalf, Charlotte Danciu and Ross filed suit on April 5 contesting the Wellington election.

The candidates' legal team was convinced that independent computer experts could ascertain whether something had gone wrong inside those ballot machines. The experts, however, would never get the chance.

Here's the thing about the ensuing discussion of having the machines print voting receipts:

• First of all, a machine could give you a receipt for how you really voted but send different information for tabulation. That is, you don't know that your receipt is telling you the truth.
• Secondly, if only the receipt is hard-copy, it means the real voting information, being purely digital, can fall apart in transmission; collecting receipts after the fact is a much more complicated process than simply hand-counting hard-copy ballots that are already on-site.
• Thirdly, receipts make it too easy to violate the privacy of the voting booth, which can potentially lead to problems like having a parent or spouse (or roommates, classmates, drinking buddies, or whoever) demand that you prove you voted the way they think you should have. Worse, it leads to potential nightmare scenarios of party thugs waiting for you as you leave your polling place and demanding that you prove you voted "correctly". (I mean, you know, we've already been told that saying you want to see someone besides Bush in the White House is "treason" - just how far-fetched is that scenario, given the riot to prevent ballot-counting by Republican operatives in the last election?)

The ballot itself is what matters. Don't get distracted with talk about receipts.