Copyright 2004 The Chronicle of Higher Education 

The Chronicle of Higher Education

September 17, 2004, Friday


LENGTH: 617 words

HEADLINE: Political Scientists Convene to Probe and Predict U.S. Elections





Labor Day has come and gone, and the U.S. election season has officially begun. A few tentative insights into the American electoral dance were offered at the annual meeting of the American Political Science Association, held here September 1-5.


Several scholars reported that recent structural reforms intended to enhance turnout on Election Day have had disappointing results. Oregon's vote-by-mail system, established in 1998, does not seem to have improved overall voting rates, said Paul Gronke, an associate professor of political science at Reed College. At best, he said, the Oregon system makes fairly regular voters more likely to vote in minor, off-year elections.


Michael J. Hanmer, of Georgetown University, offered further bad news: "Same-day registration" systems, in which people need not register before Election Day, do not increase turnout as much as political scientists had once believed. The earlier optimism, Mr. Hanmer said, was based on the adoption of same-day registration in the 1970s in states like Minnesota and Wisconsin, which already had cultures of high political participation. States that adopted the system more recently have seen much weaker gains, he said.


On another front, two scholars offered new fodder for one of the great barstool debates of our time: Exactly how much did Ralph Nader's campaign damage Al Gore's position in Florida in 2000? Michael C. Herron, an associate professor of government at Dartmouth College, and Jeffrey B. Lewis, an assistant professor of political science at the University of California at Los Angeles, estimated that 61 percent of Nader voters in Florida would have voted for Mr. Gore had Mr. Nader not been in the race.  Mr. Nader's votes did indeed spoil the race for Mr. Gore because the Florida race was extremely close -- but Mr. Nader's supporters were not overwhelmingly liberal Democrats, as many commentators have assumed. "The key thing about 61 is that it's a lot closer to 50 than it is to 100," Mr. Herron said.  Mr. Herron and Mr. Lewis based their conclusion on a study of more than three million images of actual ballots from the 2000 election in 10 Florida counties. The two scholars have analyzed "down ballot" votes -- that is, the choices that people who voted for Mr. Nader or Pat Buchanan for president made in races further down the ballot, such as those for the U.S. Senate, the state legislature, and local offices.


Another scholar looked at a more fundamental question: Why vote at all? The odds that one's vote will tip the balance are infinitesimally small. James H. Fowler, an assistant professor of political science at the University of California at Davis, proposed a new solution to this puzzle. His answer: Altruistic people are more likely to vote.  Mr. Fowler recruited 249 students to play the "dictator game," a laboratory technique pioneered by experimental economists. In the game, a player is surprised with a gift of a certain amount of money, and then asked if he would like to share a fraction of the money with an anonymous stranger. Mr. Fowler found that participants who behaved altruistically in the dictator game and who identified strongly with a political party were more likely than their egoistic peers to report having voted in the March 2004 primary elections in California.


Finally, the question on everyone's lips: Who will win the U.S presidential race? At a panel dedicated to the arcane skill of forecasting elections through the use of quantitative models, six out of seven scholars predicted that President Bush would win a majority of the popular vote in November, and even the seventh model predicted that he would win 49.9 percent of the vote.