Predicting presidential election by the numbers
It is becoming harder to predict who the Republican Party will choose to oppose President Obama. Romney was the polls' favorite just a few weeks ago, now it appears to be Gingrich. Who knows who will emerge from upcoming primaries?
Instead of trying to guess who will face the president, what is the likelihood that President Obama will be a one-term commander in chief? There is an economist who can provide that answer.
Ray Fair, a professor at Yale University, began in 1978 to use macroeconomic variables to predict the outcomes of presidential elections. What is striking is that with a relatively simple equation, one with few moving parts, he is able to accurately predict election outcomes.
Fair has boiled down the possible economic factors to a handful. The list includes the growth of output (real GDP) during the nine months preceding the election, how inflation has behaved during the president's term, and whether the economy is expanding slower, faster or at its potential.
If the economy is not expanding fast enough, this decreases the incumbent's chances of re-election. If inflation is raging, this too lowers the chances of another four years in the White House. Such a combination was enough to make predicting Jimmy Carter's one-term presidency easy.
Using these three economic measures, Fair has used his statistical model to forecast the incumbent's share of the popular vote in every election since 1960. His forecasts are quite accurate. Across 13 elections, Fair's model predicts the winner in 10 elections (including the model's accurate prediction of Al Gore as winning the popular vote in 2000). Candidate Obama's victory over John McCain was called by the model. Which elections did the model miss? It picked Nixon over Kennedy in 1960, Humphrey over Nixon in 1964, and Bush over Clinton in 1992.
Remember the Clinton campaign's mantra in 1992? "It's the Economy Stupid." The drumbeat was that Bush administration policy had failed to grow the economy and lower unemployment rates after the 1991 recession. It turns out that real GDP was expanding rapidly, at more than 4 percent in the months preceding the election. But, because these data weren't known by the time of Election Day -- GDP data are released with a considerable lag -- and with unemployment still higher than 7 percent, voters believed that Bush policies had failed and Clinton's would deliver prosperity.
Want to handicap the upcoming race? Fair's predictive model is available at his website . Plugging in your guesses at the three economic measures and you too can opine on the likely outcome of the race. Here are my predictions.
Scenario one: Status quo, meaning that real GDP grows at 2.5 percent for the first nine months of 2012, inflation stays around 2 percent and no significant drop in the unemployment rate. In this scenario, the president's predicted vote share is 49.75 percent.
Scenario two: The economy expands at a faster rate, say 3.5 percent. With inflation unchanged and unemployment starting to fall, the president garners 50.42 percent of the popular vote.
Scenario three: The economic outlook worsens due to unforeseen events (think sky-rocketing gas prices, a recession across the European Union, etc.) In this scenario real GDP growth slows to 1.5 percent over the coming months. With inflation unchanged, the model still gives the president a 49.08 vote share.
This exercise shows that regardless of whom the Republicans anoint, it will be a tight race. And if the economy has anything to say about it, improvements in economic growth, continued stable inflation and a further drop in the unemployment rate will make the incumbent hard to beat.
R.W. Hafer is a research professor of economics and finance at Southern Illinois University Edwardsville and a research fellow at the Show-Me Institute. To reach him, contact Beacon features and commentary editor Donna Korando.