Some of the biggest supporters are the disabled, who welcome the machines because they come outfitted with a variety of devices that allow nearly everyone to vote unassisted, and more importantly, in private.
James Gashel, executive director of strategic initiatives at the National Federation for the Blind, testified in favor of touch-screen machines in court last month in defense of the Maryland state elections board. "If we're going to have accessible paper trails, then they need to have accessible information for blind people as well," said Gashel, a resident of Baltimore who said that he has never had the ability to vote without help from someone else.
On the other side are groups representing minorities, arguing that touch-screen voting without a paper trail could make it easier for a corrupt elections official to make those votes disappear. "We are concerned about the fact that this is not a tried-and-true system," said Melanie Campbell, executive director of the National Coalition on Black Civic Participation. "What we are concerned about is voter confidence."
In the University of Utah survey, black voters' attitudes about voting technology diverged significantly from whites', with 52.2 percent favoring electronic methods, while putting optical-scan systems below punch cards. They were also more concerned about unintentional failures than fraud.
Some of the companies whose machines will be used in the Washington area said that they are developing versions of their machines that can produce paper records that voters can see.
"We are going to provide whatever our customers need," said Michelle Shafer, spokeswoman for Austin, Texas-based Hart InterCivic, whose eSlate machine is used by Alexandria.
Alfie Charles, spokesman for Oakland, Calif.-based Sequoia, said its AVC Edge machines in Nevada already are being used in early voting with paper trails and that the results have been good so far. The printers, he said, add an extra $800 onto the cost of a $3,200 machine.
Diebold spokesman David Bear said that the company is working on prototypes, but added that it is a solution for a problem that has not been proven. "There is no entry point for someone to hack the system," he said. "It's like saying you could hack into my clock radio."
All the wrangling that has taken place over electronic voting the past year has pointed out one inescapable conclusion, many observers say: Touch-screen machines may not be the ideal solution to voting error and fraud, but their weaknesses are simply the latest variation in a process that has never been completely secure.
"I think most Americans are very naive about the potential for fraud," said Rep. Vernon Ehlers (R-Mich.), who authored a bill calling for federal standards to govern how electronic voting machines are used. "They think that went out with the Pendergast machine and Tammany Hall, but with computers, if you get the right person involved, there's always a possibility of fraud."
Gashel noted that older voting systems have experienced problems of their own. "We're not convinced that [touch-screen voting] is a problem to the extent that it is with some other systems," he said. "I know that computers can make mistakes but mechanical mechanisms can make mistakes too."
Then there is the human factor.
"We have about 8,000 election officials in the country who'll be supervising 193,000 polling places and about 1.5 million Americans who are going to serve their fellow citizens as poll workers," said Paul DeGregorio, one of the commissioners on the Elections Assistance Commission. "We're still dealing with human beings who run elections."
R. Michael Alvarez, a California Institute of Technology professor and co-director of an electronic voting research project run by Caltech and the Massachusetts Institute of Technology, said 4 million to 6 million votes were lost nationally in the 2000 presidential election because of a variety of problems. According to a 2001 report issued by the Caltech-MIT group, as many as 2 million of those votes were lost because of faulty ballots or equipment problems, but up to 3 million were lost due to registration mix-ups, with another possible 1 million votes lost through botched polling-place operations.
Alvarez, who co-authored the Utah study with Hall, said electronic voting could cause problems, but it could also prove to be more accurate than any other voting method. "We're going to be running a massive experiment on Nov. 2," he said.
Linda Schade, of course, does not want to participate in this kind of experiment, especially one that will be done live. On Election Day, she says she'll send her TrueVotemd troops out in a statewide pollwatching effort. "You're talking about a crisis in voter confidence, a crisis in accountability in terms of the certainty of elections results," she said. "That's a big problem."
washingtonpost.com staff writer Bob Greiner contributed to this article.