Touch Screens Show High Rate of Unrecorded Votes for President in 2008
Paper Ballots Found More Efficient at Recording Voters' Choices
June 26, 2009 - A professor's study of North Carolina's 2008 Presidential election shows that optically scanned paper ballots were better at registering the intent of the voters than touch screen voting machines.
Mark Lindeman, an assistant professor of political science at Bard College in New York, found that in the 67 North Carolina counties where the voting method is optically scanned paper ballots, 0.78% of ballots failed to register a vote for President last November. The 24 counties where touch screens were the principal method of voting saw 1.36% of ballots fail to register a vote for President, a difference of over 7000 votes in the 2008 election.
"The evidence available to me indicates that in fact, optically scanned paper ballots fared better than DREs [touch screens] in recording and tabulating voter intent," Prof. Lindeman wrote.
Lindeman also analyzed demographic differences among the counties that might explain the higher number of unrecorded votes in the counties that used touch screens. He found, in fact, that paper ballot counties measured higher in factors such as less education and poverty that would be expected lead to a high rate of unrecorded votes, meaning that the "effect " of touch screens on the unrecorded vote rate was even greater than the raw numbers suggest.
Voting experts believe that a small number of voters, usually less than 1%, decide deliberately not to cast a vote for President, but that if the number of ballots that show no vote for President is higher with a given voting technology, it is a sign that the technology was less easy for voters to use, or may not have functioned properly. The percentage of ballots that fail to register a vote for a given office is called the "residual vote rate."
"DRE boosters say the residual vote rate should be lower on touch screens than on scanned paper ballots, but the performance doesn't match the promises" said Lindeman.
Prof. Lindeman's findings are consistent with previous studies showing that precinct-based paper ballot scanners have a lower residual vote than touch screen machines. A study of the Brennan Center for Justice showed that precinct-based optical scanners had the lowest residual vote rate of any type of technology in the 2004 Presidential election. In 2006, Iowa's election results for all contested statewide races showed a consistently higher residual vote rate for touch screens than for optically scanned paper ballots.
"Optical scan has a strong track record, and these findings just make it stronger," said Pamela Smith of the Verified Voting Foundation. "This is why we fought so hard for optical scan back in 2005 and 2006," said Joyce McCloy, director of the North Carolina Coalition for Verified Voting. "It turns out that the lower-tech way best serves the voters," McCloy added.
The Machinery of Democracy: Voting System Usability,” p. 5. http://brennan.3cdn.net/bb59042f6839b7fee2_njm6bcl84.pdf
“Residual Votes in Iowa November 2006,”