Problems with e-voting? Blame the humans
IDG News Service 10/19/04

Paul Roberts, IDG News Service, Boston Bureau

Voters worried that an electronic voting machine might accidentally eat their vote on Nov. 2 would be better off pointing the finger of blame at clueless poll workers than at shiny new touchscreen machines, according to information released by the Information Technology Association of America (ITAA).
The technology industry group, which is a staunch supporter of electronic voting technology, made that argument in a document that was distributed to "help journalists put election equipment-related snafus in context." While electronic voting machines aren't blameless, but are just one part of an election day puzzle that includes local election officials and volunteers, some of whom have scant or uneven training, the ITAA said.

The document outlines a number of woeful election day scenarios, from missing voter registrations to incorrect ballot information and nonfunctioning electronic voting machines. In each case, the ITAA offers possible explanations for the mishap, often with the effect of exonerating the voting machine.

For example, human error may explain voting machines that will not turn on or operate, the group wrote. Poll workers may not have plugged in the machines, or forgotten a password required to start the machines, resulting in "lengthy delays as this information is retrieved from other sources," the paper said.

The idea behind the media primer is to get journalists to better understand how electronic voting technology works and not always assume that problems with voting are due to failures of electronic voting technology, said Bob Cohen, senior vice president of the ITAA.

As examples, Cohen pointed to news stories that appeared Monday after early voting commenced in Florida.

Newspaper headlines about problems with electronic voting technology belied the real problem: There weren't enough phone lines between polling places and a central database at election headquarters that contained voter registration information, he said.

"There's a disconnect between the headlines and the actuality, which is that the machines are working pretty well, but that they're part of a process that involves people and procedures," he said.

Voter registration systems "are different from elections systems and problems with the former do not indicate problems with the latter," the ITAA said in its paper.

Not necessarily true, said Bev Harris, executive director of BlackBoxVoting.org, who noted that ES&S, a major vendor of direct recording electronic (DRE) voting systems also makes registration software for use with its systems.

Election systems vendors have an obligation to make sure localities know what is required to effectively deploy and use their systems, she said.

"If you're a manufacturer and you're making a registration system that interacts with your (voting) systems, you have to know what's required," she said.

Vendors shouldn't get a pass on mishaps by elections officials that affect the performance of DRE systems, Harris said.

"(Phone access) is part of the specification. It doesn't give them a pass. The systems should have been tested," she said.

Whether poll workers, electrical storms or internal malfunctions are to blame for voting mishaps is besides the point, Harris said.

"What matters is the result: Can (the DRE machine) cast a vote? ... Usability is partly about looking at the target group that uses it and making sure it's dummy proof I hate to use that word but (DRE machines) need to be robust enough and designed in such a way that the target group won't have a lot of problems using it," she said.

Cohen disagreed, saying that voting machines are a solution to the debacle in 2000, when punch card ballot machines obscured the results of the vote in Florida.

"These systems are secure, they're accurate, they enfranchise the physically disabled and people for whom English is not a first language they're going to make wholesale improvements to the way things are done," he said.

The truth about DRE machines is probably somewhere in between the positions taken by people like Harris and Cohen, said Dan Seligson, editor of Electionline.org, the Web site of a non-partisan Washington, D.C., group that tracks election reform across the U.S.

"There are two sides of this issue: Elections officials say that (DRE machines) are one hundred percent safe and accurate, and on the other side, computer scientists say they're fraught with problems. The truth is in the middle. No system is 100 percent secure, nor are they rife with security breaches."

The real test of DRE machines will come on Nov. 2, when some 30 percent of U.S. voters will use them to cast a vote in the presidential election.

"How (DRE machines) perform will determine whether voters continue to express confidence in them, despite what they've been reading for the past year and a half," Seligson said.