Predicting Election Outcomes
There's a curious but robust feature of US presidential elections: when the economy isn't doing well, the incumbent party usually loses.It doesn't matter which party is in power, they usually lose. You can see that now in the online odds for McCain, which have dropped a lot as bad news comes out about the economy. You can talk all you want about other things, but the economy is likely the primary driver here.
The reason I say "curious" is that there's very little evidence that presidents (or the government more generally) have much impact on the current state of the economy. They can influence long-term performance, but in the short term, the US economy, and others, go through periodic ups and downs that we call business cycles. That was true 200 years ago and remains true today. There's some debate about where business cycles come from, but the president's actions are rarely on the list. Nevertheless, it's become good strategy to claim credit for good news and blame the other guys for bad news -- even though neither is likely.
Related link: the idea comes from a series of papers by Yale economist Ray Fair. A nice summary is his 1996 piece.