Al Gore and George W. Bush

To paraphrase legendary Green Bay Packer coach Vince Lombardi, “the economy isn’t everything, it’s the only thing.” At least that’s according to Yale economics professor Ray C. Fair. He’s built a model to predict the presidential election that’s based largely on just three economic issues: Is the economy growing? Is inflation low? And how good has the economy been over the past four years?

If the answers are yes, yes and not bad, then for the election of 2012, Fair’s model would predict a clear victory for President Barack Obama. If the answers are no, no, and not good, then his model would predict a clear victory for former governor Mitt Romney.

What’s the model saying now, just days before the election?

For one thing, the economic variables are all set in stone so, unlike what the candidates are saying, they’re not changing before the election. Here they are. The economy is growing at about 1 percent per capita. Inflation is about 1.6 percent. And, Obama has had one quarter when economic growth was relatively strong. These numbers are very “so-so,” and using them, the model predicts that the president will get about 49 percent of the popular vote. In a two-party race, 49 percent for Obama, of course, means 51 percent for Romney. In other words, a Romney victory.

There’s one caveat to consider. The predicted 49.03 percent share of vote for the president is well within the “margin of error” inherent in these models. Economists, statisticians, and people who build this type of econometric model for a living, would say that even though the model predicts a Romney victory, the model’s prediction is “within the standard error” of the estimate. Translation? The actual vote share is as likely to be 49 percent for Romney as it is for Obama.

That said, Obama would have clear victory in sight if the economy were in better shape than it now is. If the economy were growing at 3.3 percent per capita in this year, which is roughly what many economists were predicting would be the case back in 2010, Fair’s model would predict Obama to receive 53.5 percent of the vote. In other words, an Obama landslide.

Although the main economic variables work against Obama, some non-economic variables work in Obama’s favor. One of these is the incumbency effect. As Fair has explained, voters tend to vote against a party that has been in power for two terms. Since the Democrats have only been in power for one term, this factor goes in Obama’s favor. The incumbent is also generally favored to be reelected, giving another point for Obama.

On the other hand, there’s a historical bias in favor of Republicans, so chalk this one up for Romney.

One way to gauge the usefulness of an economic model is to see how well it’s done in the past. On this score, Fair’s model does very well. In the 24 elections since 1916, it’s been right all but two times. It missed the 1960 election, when the model predicted that Richard Nixon would get 51.1 percent of the vote, but he received only 49.9 percent. It also missed the 1992 election, predicting that president George H.W. Bush would obtain 50.9 percent of the vote, while he received just 46.5 percent and, of course, lost his bid for reelection. (Fair’s model is based on a two-candidate election, and the presence of third party candidate, Ross Perot, threw off the estimates that year.)

Fair’s model has been right even when it’s been ‘wrong.’ His model predicted that Al Gore would receive 50.8 percent of the two party vote in 2000; he actually received 50.3 percent.

Of course, despite winning the popular vote, Gore didn’t become president. Thirty-six days after the election, Texas Gov. George W. Bush became President Bush by five electoral votes.

Will history repeat this year? Perhaps.

If the economy’s all that matters, then Romney’s set to win the popular vote. But, that’s not all that matters. For Gov. Romney to become President Romney he must reach 270 electoral votes. And, that’s no sure thing.