Political Scientists Predict Victory

Political Scientists Predict Victory for Obama, or a Tight Race
By David Glenn, Chronicle of Higher Education, September 2, 2008

Will Barack Obama cruise to victory in November -- or will the contest be Bush-Gore tight? Both scenarios were sketched here on Friday during the annual meeting of the American Political Science Association when eight scholars offered stylized forecasting models for this fall's election.

Seven of the scholars predicted popular-vote victories for Mr. Obama, but two of them forecast margins so thin that they said he might easily lose the Electoral College. The eighth panelist was not ready to make an official call because his model is based partly on Labor Day opinion polls. But he said it was very possible that his model would predict a victory for John McCain.

The panel featured its share of partisan banter and baseball analogies -- but the scholars said their forecasting models were not just amusements. As they are refined over time, they said, the models should help shed light on how democracies operate. (The object here, by the way, is to correctly predict each candidate's share of the votes cast for the two major parties. Minor-party votes are excluded from the
analyses, so the predicted shares for Mr. McCain and Mr. Obama always add up to 100 percent.)

Without further ado, in descending order of good news for Mr. Obama's camp, here are the results the analyses project:

• Brad Lockerbie, a professor of political science at East Carolina University, said his model predicted an easy victory for Mr. Obama, who would take 58 percent of the vote. Mr. Lockerbie's model is based partly on the University of Michigan's survey of consumer expectations, an index that recently registered its lowest level since 1980. "I'm less than happy with this forecast," Mr. Lockerbie said, explaining that the 58-percent figure seemed to him intuitively too high.

• Thomas M. Holbrook, a professor of government at the University of Wisconsin at Milwaukee, predicted that Mr. Obama would win 55.5 percent of the vote. Mr. Holbrook's model is based on approval ratings for the incumbent president, economic growth during the second quarter of the year, and survey data about personal finances, among other factors.

• Alan I. Abramowitz, a professor of political science at Emory University, predicted that Mr. Obama would win 54.3 percent of the vote. His model is based on economic growth during the second quarter, the incumbent president's approval rating, and the number of terms the incumbent party has held the White House.

• Christopher Wlezien, a professor of political science at Temple University, predicted that Mr. Obama would win 52.2 percent of the vote. His model is based on the leading economic indicators during the second quarter of the year, personal-income growth, approval ratings for the incumbent, and current opinion polling in the McCain-Obama race. (Mr. Wlezien will continue to update his forecast as
the McCain-Obama polls change.)

• Alfred G. Cuzán, a professor of government at the University of West Florida, and Charles M. Bundrick, a professor emeritus of statistics there, predicted that Mr. Obama would win 51.9 percent of the vote. (Their forecast was presented in absentia because they stayed home to help their campus prepare for a possible hurricane strike.) Their model is based on economic growth, the length of time the incumbent party has held the White House, and whether the incumbent party has pursued an
expansionary fiscal policy.

• Helmut Norpoth, a professor of political science at the State University of New York at Stony Brook, forecast that Mr. Obama would win 50.1 percent of the vote. Mr. Norpoth's model is based partly on how much resistance each nominee faced during the primary season. He pointed out that elections like this one, with no incumbent on the ballot, are historically much closer, on average, than elections with an incumbent. (Think of 1960, 1968, and 2000.)

• Michael S. Lewis-Beck, a professor of political science at the University of Iowa, offered a mixed message. The model he typically uses -- which is based partly on economic growth and approval ratings for the incumbent -- gives Mr. Obama a clear victory, with 56.6 percent of the vote. But he said that racial prejudice was likely to undercut that margin. Working with Charles Tien, an associate professor of political science at the City University of New York's Hunter College, Mr. Lewis-Beck used four different methods to estimate the number of voters who would be unwilling to vote for any black candidate. After completing those analyses, Mr. Lewis-Beck and Mr. Tien were left with a tiny margin of victory for Mr. Obama, with 50.07 percent of the vote. (Other scholars on the panel agreed that race was an unknown and probably volatile factor this year. But Mr. Abramowitz scorned
Mr. Lewis-Beck's procedures: "I don't make ad hoc adjustments to my model," he said.)

• James E. Campbell, a professor of political science at the State University of New York at Buffalo, uses a simple model based on Labor Day horse-race polls and economic growth during the second quarter of the year. If no incumbent is running, as is the case this year, Mr. Campbell reduces the weight of the economic variable by half. His assumption is that voters will credit or blame John McCain for economic conditions, but only half as strongly as they would credit or blame the incumbent, George W. Bush, if he were running himself. (Some of the other forecasters' models also make adjustments for non–incumbent election years.)

Mr. Campbell said he was not ready to make a prediction before Labor Day. But he said that second-quarter growth was so strong that his model would predict a victory for John McCain unless he polled less than 47 percent on Labor Day.*

An audience member said he was skeptical of the U.S. Commerce Department's new official estimate of second-quarter economic growth, which was revised sharply upward last week, to 3.3 percent from 1.9 percent. That figure seems out of sync with dire recent reports about personal income and housing foreclosures, he said. If the economic-growth figure is indeed wrong, that would cause problems for at least three of the models presented on Friday.

The forecasts will be published in the October issue of PS: Political Science & Politics. For those keeping score at home: In 2004, seven scholars -- essentially the same cast of characters -- offered forecasts at the political-science meeting. Six of the seven correctly predicted that Mr. Bush would win, and four of their models were within 2.5 percentage points of Mr. Bush's actual vote share. But in 2000 they whiffed, predicting that Al Gore would win between 52.8 percent and 60.3 percent of the two-party vote. His actual share was 50.2 percent.

*Note: In past elections, Campbell has used Labor Day polls. Historically, both parties’ conventions are held before Labor Day. This year, the Republican convention is scheduled to conclude several days after Labor Day. The average of five national polls August 29 - Sept. 1 show Obama leading 49.0% to 43.0%. That is, just before Labor Day, McCain claims 46.7% of the decided two-party vote. These polls,
however, were done after the Democratic convention and before the Republican convention. Very likely, polls conducted after the Republican convention will show McCain polling higher than 47%.