Study Finds Millions of Votes Lost
Universities Urge Better Technology, Ballot Procedures
By Guy Gugliotta
Washington Post Staff Writer
Tuesday, July 17, 2001; Page A01
Between 4 million and 6 million Americans either failed to cast votes or had their votes invalidated in last year's presidential election because of faulty equipment, mismarked ballots, polling place failures and foul-ups with registration or absentee voting, a study by two leading universities reported yesterday.
The study recommended that election boards get rid of punch cards and lever machines, curb absentee voting, institute registration reforms and discard any consideration of Internet voting for the foreseeable future. It urged the federal government to help fund equipment upgrades, to create a new agency to set up election technology standards and to conduct research into improved electronic voting systems.
"The main conclusion is that there has been precious little study of these issues and precious few resources" applied to resolve them, said California Institute of Technology President David Baltimore. "The voting process has previously simply not been taken seriously."
The Voting Technology Project report, by the California Institute of Technology and the Massachusetts Institute of Technology, has been eagerly anticipated by both elected officials and other would-be reformers as a nonpartisan and carefully researched scientific study of the mechanics -- and not the politics -- of holding an election.
The report said that in presidential elections, punch card ballots, like those that caused last year's Florida debacle -- have an error rate of 2.5 percent, the worst of any method used. Lever machines, while maintaining a relatively low rate of error for presidential voting -- 1.5 percent -- were the worst performers when voters cast ballots for lower offices. The incidence of invalidated ballots rose to 7.6 percent for governor or senator.
But while the university recommendations received a generally warm reception, several officials and organizations expressed reservations about the report's emphasis on technology and its promotion of a federal role in a process that traditionally has been controlled by state and local governments.
In Little Rock yesterday, the National Association of Secretaries of State presented a summary of the "best practices" used in member states, suggesting that election officials consider a broad spectrum of reforms. Secretaries of state, whether elected or appointed, oversee elections in most states.
The NASS recommendations overlapped many of those of Caltech and MIT, but they focused more on voter education and the training of election officials. "I hope people aren't convinced that simply plugging in new voting systems around the country is going to solve our problems," said Kansas Secretary of State Ron Thornburgh, the newly elected NASS president.
Congress has several competing election reform bills pending, and many of them accept a federal role in the process. In the House, Reps. Robert W. Ney (R-Ohio) and Steny H. Hoyer (D-Md.), the chairman and ranking member of the House Administration Committee, respectively, plan a bipartisan bill that will eliminate punch cards and provide assistance to states to upgrade equipment.
"This study backs up what Mr. Ney and I have contended -- the federal government must play a role in replacing antiquated election systems as part of the effort to ensure that every vote cast is a vote counted," Hoyer said. But he noted that a cornerstone of the legislation will be that "the federal government should not mandate solutions."
Caltech and MIT, funded by the Carnegie Corporation, announced the voting project last December, with Baltimore suggesting that since the Florida debacle was largely a technological mishap, technology specialists ought to be called upon to fix it.
The teams did not consider political or sociological issues -- such as the high rate of invalidated ballots in minority precincts. But the two university teams, comprised of political scientists and technologists, went far beyond a simple machinery analysis to look at every aspect of election procedures.
The teams used figures from the U.S. Census Bureau's Current Population Survey, which showed that 40 million registered voters in the United States did not participate in the November 2000 elections.
Roughly 7 percent -- or about 3 million voters -- said they didn't vote because of registration problems, and another 1 million cited "long lines" or other polling place shortcomings. Researchers said the registration foul-ups could be lower or higher.
And using election returns data, the teams estimated that faulty equipment or confusing ballots caused 1.5 million to 2 million votes to be unmarked or mismarked. In all, the team estimated that between 4 million and 6 million votes were lost in the election.
"There are failures at every step," said MIT political scientist Stephen Ansolabehere. "That's an unacceptably high rate of vote loss, and by 2004, the United States can cut the number of lost votes in half using existing technology."
The study recommended getting rid of punch cards, used by one-third of America's voters, and lever machines, used by 17.8 percent of voters. It also recommended that elections boards discard the more primitive electronic on-screen machines because they are confusing and slow-moving. About 11 percent of voters used electronic ballots in 2000.
To replace these technologies, the report suggested a move to optical-scan ballots, like the answer sheets in college board exams, and research on new forms of electronic voting. The report emphasized that optical-scan ballots should be authenticated at the precinct level, so voters will know when they have made a mistake and be able to correct it.
The report laid great emphasis on the need to reform registration processes and polling place techniques. It suggested giving election officials computer access to the complete voter rolls at the polling place to facilitate checks, and proposed that states develop qualified voter files.
The report was harshly critical of absentee voting, saying it reopened the electoral system to coercion, a popular Election Day ploy during the 19th century, and did not increase voter turnout. The report suggested that on-demand absentee voting be "restricted or abolished."
The teams expressed even less enthusiasm for Internet voting, which "has all the problems of absentee voting and adds problems of security," said MIT computer scientist Ron Rivest. "At least a decade of further research on the security of home computers is needed before Internet voting can come in."
Staff writer Shankar Vedantam contributed to this report. |
