Between 4 million and 6 million ballots were lost during the US's contentious 2000 elections, in good part due to insufficient and "make-do" technology, a group of scientists and engineers reported Monday upon the release of their seven-month study on US. voting.
The study, entitled "Voting: What is, what could be," is part of an overall project undertaken by a team of 10 computer scientists, mechanical engineers and political scientists from the Massachusetts Institute of Technology (MIT) and the California Institute of Technology (Caltech) to tackle some of the issues faced during last year's election debacle in Florida.
"Millions of voters were disenfranchised by the complications inherent in present technology," said Caltech President David Baltimore at a press conference Monday.
Of the votes erroneously cast aside during last year's elections, 1.5 million to 2 million were lost due to faulty equipment or confusing ballots, while 1.5 million to 3 million were squandered on registration problems, the study indicated.
Baltimore and MIT President Charles M. Vest wrote of their disdain at the figures in the prologue of the report, stating, "in the last election, Americans learned that at the heart of their democratic process, their 'can-do' spirit has 'make-do' technology as its central element."
Of the technologies currently used to conduct voting in the US., the group found that the most unreliable are punch cards and lever machines. According to the study, punch cards had the highest rate of unmarked, uncounted and spoiled ballots over the last four presidential elections. Lever machines had the worst record with these problems during elections for Senate and governor over the last 12 years.
"Both punch cards and lever machines should be done away with," the report states.
Ironically, the low-tech hand-counted paper ballots and optically scanned paper ballots were the more reliable methods.
"The challenge is finding technology that is easier to use and more familiar than paper - we are not there yet," said MIT Political Science Professor Stephen Ansolabehere.
Ansolabehere noted that part of the problem is that the industry that produces voting equipment is very small, gaining revenues of between just $US150 million and $200 million a year, which puts them in a difficult position to do costly research and development.
The team is suggesting that the federal government create and fund a system for evaluating equipment in order to develop the most effective equipment, without having to rely on the companies that sell it.
And given that significant research is needed, high-tech solutions may not be the answer in the near term. The group cast a dubious glance on touch-screen ATM-like voting machines, saying that they have not been sufficiently tested to measure whether the systems are secure and the user interfaces are easy-to-use.
The group also recommended that Internet voting be delayed, given that the security of user PCs cannot be confirmed. Furthermore, Internet voting brings about some of the same concerns stirred by absentee voting - that voter cohesion, a lack of privacy and possible manipulation of ballots would be more prevalent when voting outside of the booths, the group said.
And although touch-screen and Internet voting may be a while off, short-term solutions, like replacing punch card machines with optically scanned machines can provide at least a short-term fix, the group suggested.
If changes are not made, warned Ansolabehere, "we are at risk of another controversial, contested election like the one we had in Florida."
To find out more about the study, and the group's recommendations, visit http://web.mit.edu/voting.