So which side is right? Well, for what it's worth, the current betting on Intrade has Obama's odds of being reelected at 61 percent. But we'd add a few more points to the mix. First, there's always an inherent bias among the press and political classes toward predicting close elections. Reporters want to cover an exciting race, and campaigns want their supporters energized to turn out. And of course, no one actually knows what will happen, so saying it will be close is always a safer bet.
Second, among those who are predicting a blowout victory, virtually no one – with the recent exception of conservative commentator Dick Morris – is anticipating that victor to be Romney. (As we've written before, even many Republicans are not particularly bullish on Romney's chances.) By contrast, back in 2004 – a campaign that strikes many as similar to this year's, only with the parties playing reverse roles – there were a number of prominent "Kerry landslide" predictions.
One argument often cited by the "landslide" camp, then as now, is that history shows incumbents tend to win or lose reelection by large, not small, margins. In 1980, Jimmy Carter lost to Ronald Reagan 49 to 489 in the Electoral College; in 1984, Reagan beat Walter Mondale 525 to 13; in 1992, George H.W. Bush lost to Bill Clinton 168 to 370; and in 1996, Clinton beat Bob Dole 379 to 159.