Here is an interesting counterpoint to Allan Lichtman's "infallible" presidential political prediction formula. As the author notes, "The problem with models is, they often work really well until they don't.":
This article is making the rounds this morning, suggesting that Obama's got a lock on 2012. According to a model designed by American University professor Allan Lichtman, which has successfully predicted every election since Reagan, Obama has enough of its thirteen "keys" to ensure that he will retain the White House in 2012. I thought it would be interesting to see how Hoover stacked up on the same model.
(Comparison of Hoover and Obama on Lichtman's key points omitted.)
...
Yet Hoover didn't just barely lose--he was beaten like a rented mule. The problem is obvious, I think: unlike the economic models that rely on external metrics, perception is doing a lot of the work here. Do we count Obama's stimulus but not Hoover's? And if so, why? I understand you might say that Obama's was larger, more effective, etc but this is a political model--it's about perception, not the economic effect, which is captured in other variables. Hoover's deficit spending was perceived as a big deal, and became an issue in the 1932 election.
...
As Joyner notes, this model weights everything equally. I submit that even if Hoover had had a few more keys--a minor military victory similar to capturing Osama Bin Laden, a wildly charismatic personality--his electoral results would not have been noticeably different. The economy just dwarfed everything else.
Completely absent from this analysis is a discussion of why no one in Congress will challenge the president, despite his falling approval numbers.