Who will win the vote? And why?

There’s Bob Forsythe and colleagues at the Tippie College of Business’ Iowa Electronic Markets (IEM), who are conducting an on-line futures market. There are professor of political science Michael Lewis-Beck’s political economic models for forecasting election results. New on the scene is assistant professor of political science David Redlawsk’s research into how voters use campaign information to make decisions. And there’s political science professor Arthur Miller’s Heartland Poll, a unique measure of political opinion among Iowans and residents of six surrounding states.

IEM offers predictions, day-to-day analysis

Bob Forsythe

The IEM was launched in 1988 by four Iowa professors to test a hypothesis: that the free market could pick a winner more accurately than opinion polls. They turned out to be right. In the 1988 and 1992 presidential contests, the markets beat every major national poll, predicting the winners’ margins within two-tenths of a percentage point.

Forsythe, a professor of economics and senior associate dean of the Tippie College of Business, said the markets work "because they require people to put their money where their mouth is."

The IEM is a real-money political futures market operating as a not-for-profit teaching and research tool. Traders log onto www.biz.uiowa.edu/iem and bet on the outcome of a variety of issues including national elections.

As of fyi press time, contrary to current opinion polls that show George W. Bush leading in the presidential race, the IEM has Al Gore and Bush in a head-to-head race approaching the November election.

"This is a terrific education tool," Forsythe said. "The market is building up a track record inside and outside of the classroom. People know that we’ve been accurate in the past so they’re starting to look to us more."

Lewis-Beck’s model predicts winner months out

Michael Lewis-Beck

There’s another Iowa scholar garnering worldwide attention for his political predictions. Lewis-Beck is an expert in comparative politics and election forecasting. For more than 10 years he’s made a science out of the conventional wisdom in American politics that a strong economy greatly favors incumbent presidents and/or their parties at election time.

Lewis-Beck has a strong record of accuracy in forecasting elections. In the 1996 presidential election, he closely predicted the percentage of the vote that President Clinton would win.

"We beat every model, every poll." he said, "We were within one-tenth of one percentage point."

His prediction for November? A strong economy and a popular president will propel Al Gore into the White House with 55.4 percent of the two-party popular vote.

Redlawsk dissects voter decision making

David Redlawsk

Assistant professor of political science David Redlawsk brings a different perspective to public opinion—that of a former politician. Redlawsk looks at how voters use campaign information to make decisions.

"My research is predicated on the idea that better decisions are made when people have more information," he said. "But the reality of politics is it is not very important to people most of the time. They want to make accurate decisions, but they also don’t want to invest a lot of time because the pay-off seems limited to individual voters. That’s why a lot of people vote along party lines."

Redlawsk and Richard Lau at Rutgers University have developed a methodology to trace voter decision-making through a simulated presidential election environment they call The Ballot Box. This methodology was used to discover the extent to which voters could sift through information in a computer program and then ultimately vote for the candidate that best fits their needs.

His early research indicates people are more likely to vote accurately when there are fewer candidates and when campaign resources are evenly distributed between the candidates.

The long-term goal of the project is to understand how information informs the electorate and to come up with specific recommendations on how information ought to be made available to the public by politicians and political groups.

Heartland Poll asks why people vote as they do

Arthur Miller

Just down the hall from Lewis-Beck and Redlawsk, Miller is spending long hours working up election data of his own. The political science professor is director of the influential Heartland Poll and head of the University’s Iowa Social Science Institute.

Between October 1999 and January 2000, the institute conducted four separate surveys, mainly focusing on Iowa as the caucuses approached. The surveys tracked people’s perceptions of different candidates’ character, on candidate debates, and what Iowans and New Yorkers think of Hillary Clinton’s run for the Senate seat.

And the culmination of the institute’s efforts this year will come in October—just one month before the presidential election—when the group conducts its flagship Heartland Poll. The poll, which gathers opinions from Iowans and the six surrounding states, is the only one in the country that compares numbers from different states.

"We try to leave the predictions to Michael Lewis-Beck," Miller said. "We explain why people have the kinds of preferences they do and why they prefer one candidate to another and which issues they think are important to the election."

Article by Melinda Pradarelli

(This story is excerpted from one that appeared in the March 2000 issue of Illumine.)