IEM Shows How to Predict a President
Forget Iowa and New Hampshire, or 7-Eleven coffee cup polls, or astrologers. People vote their pocketbooks.
Every four years, Iowa and New Hampshire are the first states to cast votes to select the nominees for president. But does that really mean they're bellwethers of national political sentiment? In the last five contested Republican battles, voters in Iowa wrongly predicted the eventual Republican nominee three times. A coin flip would have done better. And in 2008, New Hampshire Democrats went for Hillary Rodham Clinton over Barack Obama.
So if you can't count on Iowa and New Hampshire, where should you look to get an accurate reading on the country's political future? One option might be the United Astrology Conference, to be held in New Orleans in late May. There, the group plans to divine the likely winner of the November elections from the position of the planets. In May 2008, the astrologers noted that Saturn's predicted opposition to Uranus on election day implied upheaval and transformation—the kind of "change" that would sweep Obama into the white House.
Or perhaps you'd prefer looking to market forces to predict the outcome? The University of Iowa College of Business provides an electronic futures market that allows people to place bets on real-world events, including presidential elections. The 2012 market has been trading since July 1, 2011, with "unnamed" Democratic and Republican nominees. As of now, the implied predictions are about even, with prices of 50 cents for the unnamed Democrat and 49.8 cents for the unnamed Republican. If you pick the winner, you get $1 on your 50-cent investment, but you get nothing if the candidate you put your money on loses. On the eve of the 2008 election, the prices were 90 cents for Obama and 10 cents for McCain. This market has correctly predicted every presidential election since 1980.
For those who do not want to wager real money but would like to express a preference, there is the 7-Eleven coffee cup poll. In October, 7-Eleven stores will be selling coffee in either blue (Democrat) or red (Republican) colors. Consumers pick their color and implied choice. This poll has been right since its launch in 2000.
Timing is, of course, a crucial factor in the track record of these election forecasts. Predicting the winner on the eve of an election is a lot easier than making a prediction in January. Last-minute opinion polls (including prediction markets) and voter exit polls have great track records. But their success is no more impressive than predicting the Super Bowl victor with 1 minute left to play. Getting that prediction right at the beginning of the season is a lot harder.
Timing is one of the things that make Ray Fair's predictions so impressive. The Yale economist made his predictions for the 2012 presidential race way back in October 2010, long before anyone knew who would be challenging Obama this year. In fact, professor Fair says it does not matter who the challenger is. The only thing that matters in Fair's model is the economy. Echoing Bill Clinton's mantra—"It's the economy, stupid"—Fair offers a model of total economic determination. As he sees it, people will vote their pocketbooks. Period.
Think about that proposition for a minute. Nothing else matters? Not who the Republican nominee is? Not the vice presidential nominee? Not money spent on the campaign? Not the number of candidate appearances or goofs? In Fair's model, it's all about the economy, plain and simple. And he's been right 90% of the time.
Fair's equation uses only two variables: economic growth and inflation. On the theory that voters will reward the incumbent who delivers on improving the economy, the calculation adds points for growth (as measured by changes in real gross domestic product) and subtracts them for increases in the rate of inflation.
Fair's model assumes voters have short memories. The economic despair of the last three years hardly matters. What counts, he says, is how well the economy is doing in the election year. That has to worry Republicans who want to make the prolonged downturn a central campaign issue.
In October 2010, Fair was predicting another decisive Obama victory in 2012. His prediction depended on a forecast of a strong 2012 economy. Fair assumed GDP growth of 3.69% in the first nine months of the year, including three "good news" quarters (in which GDP grows by more than 3.2%). Most other forecasters—the Conference Board, the National Assn. of Business Economics, the Congressional Budget Office, Goldman Sachs, etc.—foresee much slower growth and no "good news" quarters.
But even modest growth could make the election a lot closer than people now anticipate. So if you are handicapping the outcome, you had better keep an eye on the economy, not the candidates, the stars or the opinion polls.
Brad Schiller is a professor of economics at the University of Nevada, Reno, and the author of The Economy Today.