First there were the politicians, the pundits, and the lawyers. Now scientists are weighing in on the contested U.S. presidential election. A study shows that the controversial "butterfly ballot" used in Palm Beach County, Florida, is indeed confusing, as many (mostly Democratic) voters have argued. The paper, slated for publication in next week's issue of Nature, was released on the journal's Web site today, a press release explained, "because of its immediate topicality."
The idea to subject the ballot to scientific scrutiny arose the day after the 7 November election, says psychologist Robert Sinclair of the University of Alberta in Edmonton, Canada. The same day, in collaboration with colleagues at Pennsylvania State University, University Park, Sinclair had 324 students at the University of Alberta participate in a mock election for Canada's prime minister. Voters received either a single-column or a dual-column (butterfly) ballot, styled after the one in Palm Beach County. After they voted, the students were asked who they had intended to vote for, and to rate their confusion with the ballot format on a seven-point scale. The following day, Sinclair set up a mock polling station in a local shopping mall and repeated the test with 116 shoppers of all ages.
The students weren't easily befuddled by the butterfly ballot: Not one mistakenly voted for the wrong candidate. But of the 53 shoppers who got the butterfly ballot, four made the wrong choice, while none made a mistake on the single column ballot. Furthermore, both students and shoppers rated the butterfly ballot significantly more confusing, Sinclair says. He attributes the students' success to practice in filling out forms and multiple choice tests. "The students are real good at [that]," Sinclair says, "but people in the real world aren't."
Sinclair says he doesn't know whether the study should carry any weight in the ongoing court case over the ballot; the take-home message should be to always involve the social sciences when designing a ballot, he says. Still, other scientists say the implications are clear: "There's very little doubt Gore lost approximately 2000 votes due to the nature of the ballot," says Norman Schwartz, a psychologist currently on sabbatical from the University of Michigan, Ann Arbor.
Nature manuscript editor Rosalind Cotter insists that "we're not trying to make a political statement." Cotter says that the paper is "a fairly moderate contribution" to science, but she doesn't know whether the paper would have survived Nature's stringent review process if it hadn't been for the struggle over the presidency. "It's hard to say. It came in at just the right time."