Everyone has their theory about how the presidential election is going to play out.

Some people think Barack Obama — lately dubbed “The One” — is on his way to being anointed. Other prognosticators argue that despite the Republican Party’s woes, the nation is still essentially conservative, and will end up backing John McCain.

Some people think it all comes down to Ohio

Then there are the social scientists, who have their own ways of trying to foretell the political future.

Many of them argue for forgetting the polls suggesting that Mr. McCain is closing in on Mr. Obama. Social science suggests that given the dismal state of the economy, barring some freakish political accident the Democrats could field a cardboard box in November and still beat any Republican.

In his book Unequal Democracy published earlier this year, Princeton political scientist Larry Bartels looked at all presidential elections since 1952 and found that with one exception, in 1996, the incumbent party lost every time real income per head was growing less than 2 percent in the election year.

Unfortunately for Mr. McCain, these days income isn’t growing at all. Indeed, the economy looks so bad that Mr. Obama could win in a blowout.

Economists at Macroeconomic Advisers tweaked a model originally developed some 30 years ago by Yale economist Ray Fair to come up with a forecasting formula. It relies partly on political considerations — whether the incumbent is running for re-election, which party is in the White House and how many consecutive terms it has been in office. The other part is economic variables — the change in both real disposable personal income and the price of oil in the first three-quarters of the election year, and the unemployment rate in the third quarter.

This model correctly predicted the winning party in 12 out of the last 14 elections. Plugging in some fairly conventional economic forecasts for this year, it gives Senator Obama a ten point margin.

Or consider the model created by Alan Abramowitz, a political scientist at Emory University.

It correctly forecast the winner of all but one of the presidential elections since 1948 based on only three variables: the economic growth rate in the second quarter of the election year, the popularity of the incumbent at mid-year, and whether the president’s party has controlled the White House for two terms. By this model’s reckoning, McCain could lose almost as badly as Jimmy Carter in 1980.

So, it may be that Mr. Obama is wasting his time touring the Middle East and Europe — it may be that all he has to do is stay home and watch the nation’s weak economic fundamentals propel him to the White House.

On the other hand, it could be that academics make too much of these correlations, and the personality and backgounds of the candidates, their stands on the issues, and the dynamics of the campaign matter more.

No matter who’s right, one thing’s clear: it’s a good idea to try to win Ohio.