(FORTUNE Magazine) – President Bush ought to win this election in A landslide, and I can prove it. Yet the really interesting question is whether he can win at all.

A burst of new polling results and economic data makes the win-in-a-cakewalk scenario seemingly hard to argue with. It goes like this:

Despite the overwhelming torrent of Iraq news in recent weeks, the No. 1 issue for most voters seems to remain jobs and the economy. In the latest CBS News poll, conducted after two weeks of horrible Abu Ghraib headlines, Iraq climbed only as high as a tie with the economy when voters stated their No. 1 issue. If the Iraq news cools down just a bit--a good bet--the economy would resume its usual spot at the top of voters' concerns.

The economic news is excellent overall. The latest Labor Department report shows a huge increase in new jobs in April, following an even huger increase in March--no more jobless recovery. The economy has been expanding for 22 years and really booming since last July. Companies are investing bigtime. Inflation, despite the market's fears, remains extremely low. Retail sales have been rocketing for over a year. This is a beautiful picture.

A famously accurate predictive model confirms what these facts suggest. Professor Ray C. Fair of Yale has developed a formula that forecasts presidential elections based on economic data, and it has been extraordinarily prescient, even in the 2000 squeaker. It shows President Bush winning with a blockbuster 59% of the vote. As economic conditions change, that number could rise or fall--for example, recent worsening inflation hurts Bush's chances--but no set of even remotely plausible conditions would cause it to fall below 50%.

With the outcome apparently so far beyond doubt, where's this election's drama? The answer, obviously, is that no one has told the voters what the formula says. In the latest poll they say they'd elect Kerry by a wide margin, even with Nader in the race. Despite all that lovely economic data, they hate Bush's handling of the economy. They actually believe, by 32% to 23%, that the economy is getting worse rather than better, though the facts clearly say otherwise.

What's going on? Plenty of evidence suggests that people do indeed vote mostly on economic factors, even with major non-economic issues like Iraq, terrorism, and same-sex marriage in the headlines. But the fact that the economy is quantifiable does not mean people vote according to the economic numbers. On the contrary: As on all issues, people vote according to how they feel.

Which raises the critical question of why they wouldn't feel wonderful in light of the economic news. Unfortunately for the President, a couple of good reasons present themselves. First, even with recent knockout job growth, the U.S. is still some 1.5 million jobs short of where it was when he took office. That's a lot of people who aren't feeling optimistic no matter how much good news they hear on TV.

Second, and more important, the idea that jobs are a disaster area is the kind of notion that persists a long time no matter what. Social psychologists have done fascinating work on this general topic and have found that strongly emotional ideas--regardless of their validity--tend to spread faster and wider than merely factual ones. Hearing about someone getting fired so that his job can be sent to India is powerfully emotional. Reading about 4.2% annualized GDP growth is not.

Researchers have also identified persistent ideas as those that are easy to confirm and hard to disprove, again regardless of actual truth or falsity. Job loss is just that kind of topic. Companies lay off workers thousands at a time and issue press releases about it, but they hire workers one by one and announce nothing. The evidence available to most people thus tends to confirm the hypothesis that jobs are disappearing, even when they're not.

Professor Fair's predictive formula has suffered only one terrible failure: In 1992 it forecast that President George H.W. Bush would win reelection easily. Then, as now, America was experiencing good economic news and bad economic feeling. President Bush said, "The economy is doing well," and he was right. Bill Clinton said, "I feel your pain," and he won.

As critical as the economy's state obviously is, the voters' elusive psychological state is far more so. For deep human reasons, today's foul economic mood will be perversely slow to change, which is one reason this laboratory landslide is so far from a sure thing.