Study: Voting technology needs work; Scientists say more research is needed to prevent problems

By Ian Hoffman, Alameda Newspaper Group, September 26, 2004



Excerpts:



An all-star cast of elections scientists holed up last weekend to examine the state of voting technology and elections. It's not disastrous, they concluded, but hardly a pretty picture either.



The prescription: More research, naturally.



"There's a clear message here that there's a lot that we don't know that can get us in trouble," said Shirley Malcom, head of education for the American Association for the Advancement of Science, the world's largest scientific society and sponsor of last weekend's workshop.



If this sounds like an echo of four years ago, it is.



Scholars of American democratic practice rallied after Florida's chad-filled nightmare with a lofty goal: to rid America of bad voting technology and ambiguous elections. Congress promised tens of millions of dollars to put U.S. voters and voting technology under a microscope, then $2.8 billion to buy new voting equipment.



But the research money never flowed. And consensus on a voting system capable of accurate, verifiable and secure recording of the nation's political desires remains elusive.



------



Despite more than 200 years of honing, the tools of U.S. elections are hardly the best in the world and remain vulnerable.



"Basically, we should be serious about this like other countries," said MIT computer expert Ted Selker. "We are the most complex voting country in the world, but we are not the most sophisticated at all."



Absent systematic and federally funded research, a handful of small and mid-sized firms designed the new voting systems for the 2004 elections in secret. The most vigorous testing is by private laboratories hired by the voting-machine vendors.



Federal standards for voting machines were updated in 2002. Yet all voting machines slated for use in large numbers in November only meet 1990 standards or are grandfathered, exempt from any standard. States offer a mixed bag of requirements for additional testing and approval.



Security vulnerabilities and outright mistabulation errors slipped past certification testing and approval.



"What's still frustrating and difficult is the certification process," said Stephen Ansolabehere, an MIT political science professor and co-director of a joint MIT/CalTech voting research project . "If there was one thing the state and federal government really got distracted from that was unfortunate, it was the development of the standards and testing process."



------



Where Washington lost interest, private organizations such as the Carnegie Corp. and the Knight Foundation supplied close to $1 million a year for scientists to study elections matters. MIT's Selker said he thinks at least 10 to 20 times that amount is needed.



But where will the money come from, if a divisive, court-mediated result in 2000 and a high-stakes election in 2004 were insufficient to pry funds loose from Congress in the interim?



MIT's Ansolabehere says it may take "another big mess.".



"If things don't go well in some of the states, then there will be an added impetus," he said. And if the election goes smoothly? "It will take some of the pressure off."