Mercuri: “Poor engineering” leads to “high-tech disenfranchisement”

Long before the notorious hanging chads focused the country’s attention on the 2000 presidential race in Florida, computer security expert Rebecca Mercuri was concerned about the vulnerability of the nation’s voting methods.
On Nov. 10, Mercuri, who came to Lehigh as part of the Computer Science and Engineering Department’s Distinguished Seminar Series, shared her long-standing and well-researched concerns about “E-voting in an Untrustworthy World” with a group of students and professors.
“We don’t like to believe that we’re living in an untrustworthy world, or that our votes might not be secure, but that’s the reality of the situation,” Mercuri, a former Harvard Fellow-turned-voting-consultant, said. “It is a fact that every vote does not count, that votes vanish, and that high-tech disenfranchisement exists. I’m not saying it’s deliberate, but it is poor engineering.”
In the wake of the contentious 2000 election in Florida, Mercuri said that politicians rushed to create and vote into law a “remedy” called the Help America Vote Act (HAVA). That act was designed to update clumsy, out-dated, inefficient and arbitrary counting methods through better design of election machinery. The new equipment would allegedly streamline the process of casting, verifying, counting and recounting votes.
Problems persist
Nearly $3.8 billion in federal funds was made available, prompting election officials to rush out and purchase flawed and already outdated equipment, Mercuri said.
“They were designed to very poor standards,” she said. “(Voting machine) vendors were out there telling people they had to have a computerized system, which was not true. Paper balloting and counting, which are far safer, would satisfy the HAVA requirements, too. And most of the election officials were under the impression that this HAVA money had to be spent quickly.”
Mercuri noted that the new touchscreen voting machines and other computer-assisted voting methods are not only more likely to break down during periods of heavy use—such as on election days—but are highly vulnerable to hacking or malicious programming.
“All over the country, local governments were rushing out to buy these new black box electronic voting systems without really understanding how they are used, and how they really don’t provide what is essential in a disputed election, and that is the ability to conduct a recount,” Mercuri said. “We haven’t really fixed the problems. If anything, we’ve gone backwards.”
Look no further than the 2004 presidential election, she said.
“Unfortunately, what we had in many states where these voting machines were in place, was no way to know if ballots were cast or counted, and no way to find out if the machines tallied votes in an accurate, unbiased, independent way. We don’t even know if such a process was even possible,” Mercuri said.
“An objective assessment”
Mercuri became absorbed with the topic in 1989, when she began her research related to electronic voting issues and the sanctity of the vote. Her work involves the formulation of a methodology for striking a balance between anonymity and transparency in computational systems, concepts that are inherently intertwined and even contradictory, especially in security-related applications, but whose relationships are not well understood.
She said that she does her part by continuing to study e-voting issues, providing expert testimony on disputed elections (including Bush v. Gore in 2000), lecturing around the country, and attempting to educate election officials and the general public on the threats such poorly designed and insecure systems pose to the democratic process.
“I proselytize to the good reporters out there who are really interested in this topic,” Mercuri said. “As an engineer, I can provide an objective assessment, and pose the important questions, like why are we allowing software to be out there with secret controls?”
At this stage, she noted, only three corporations control most of the voting systems used in the United States and around the world, and these systems are plagued with problems.
“Votes go missing, machines break down,” she said. “And in the midst of all this, a few votes here and there could shift.”
She cited a Yale study, which showed that a one-vote shift in every precinct in the country could change the Electoral College outcome.
Despite the easy acceptance of flawed systems by election officials around the country and disinterested mainstream media outlets, Mercuri said that a growing number of private citizens are sounding the alarm.
“More and more, I keep running into people who really seem to understand the threats posed by e-voting,” she said in an interview earlier in the day. “It would help if the media shared that.”
If anything, news outlets exacerbate the problem by demanding election results quickly.
“They want the results for the 11 p.m. newscast,” Mercuri said. “They want them now, and they don’t want to wait it out. They do not care if the results are wrong, as long as they’re quick.”
Introducing Mercuri was Henry Baird, professor of computer science and engineering, who shares her concern for national voting methods.
“She has a passion for protecting democracy,” said Baird, who met Mercuri while they were both young scientists at Sarnoff Laboratories in Princeton nearly 25 years ago. “She was first drawn to this subject matter out of a sense of civic duty when she learned that Bucks County was contemplating the purchase of a flawed voting system. Since then, through years of study, she has established herself as one of the nation’s foremost authorities on this topic.”
--Linda Harbrecht