President George W. Bush ’68 is likely to win this year’s election, according to Yale economics professor Ray Fair’s mathematical formula for predicting presidential election results.
Fair designed his formula in the 1970s to determine election outcomes based largely on economic factors. In an update last Friday, he said his calculations suggest that Bush will receive 56.3 percent of the vote in November.
“There’s no sensible economic forecast that would lead this equation to predict that the Republicans would lose,” Fair said.
Fair’s equation relies largely on economic growth, inflation, and the number of quarters during which the rate of real GDP growth per capita was greater than 3.2 percent. It also incorporates such factors as the candidate’s incumbency status and the duration of a party’s power.
The equation’s prediction is off by an average of 2.4 percent each year, Fair said. In both 1996 and 2000, its margin of error was approximately 1 percent.
“It’s one of the oldest and best-known attempts to forecast election outcomes using mainly economic factors,” political science professor Donald Green said.
Fair said that historically, Republicans have had a slight advantage over Democrats, other things being equal. He also said the duration of a party’s power begins adversely impacting a candidate’s chances of being elected once that party has been in power for two consecutive terms — a problem which President Bush does not face this year.
“The message of it is that Bush has this big head start and the economy’s pretty good,” Fair said. “Inflation’s fine, and growth this year looks like it’s going to be okay, if not better than okay.”
In order for Bush to lose the election, Fair said there would have to be negative economic growth in the last three quarters of this year and a “substantial” increase in inflation.
Green said many political scientists, especially at Yale, have attempted to make predictions like Fair’s. Green said Fair has updated his equation in order to improve its accuracy after each election, but no formula is foolproof, particularly for close elections, he said.
“It sort of tells you on balance what would be the outcome if the economy was the sole [determining factor],” Green said. “People have often treated [predictions] as if they had some mysterious quality to them, but they’re actually fairly straightforward.”
This election will be a “strong test” for the equation because it forecasts such an overwhelming victory for Bush, Fair said.
“If Bush loses, or wins by just a little bit, it would be a very large error for the equation,” he said.
In his 2002 book, “Predicting Presidential Elections and Other Things,” Fair acknowledged possible “pitfalls” of his formula.
“I try in this book to point out the strengths and weaknesses of the method I’m using,” he said. “When people see this equation, they need to understand what its pluses and minuses are.”
Fair wrote that one potential drawback to his system is that the outcomes of previous elections may have been caused by factors not accounted for in the formula and may have only corresponded with his forecast by chance.
Fair’s Web site provides extensive information on the formula, including past predictions and periodic updates. To generate additional predictions for this year’s election, users can also enter values for economic growth rates, inflation and number of quarters of economic growth.