Why Those Election Predictors Are Useful

Brit Hume tweets (and, yes, I know; Brit Hume tweets? but that's what's out there...):

Can Barack Obama campaign his way to re-election in the face of conditions in US that would normally spell defeat for any incumbent?

This is exactly why those election prediction systems are worthwhile. Well, you don't need them, exactly; you could take postelection explanations, instead. But, basically, it's extremely useful to be able to specify what happens when presidents run for re-election "in the face of conditions" such as those Obama will face this year. The answer, of course, is complicated...but still, we have some pretty good evidence that it's the current rate of growth, not the overall level of growth over four years or the absolute level of any indicator, that seems to matter. And by that measure, what Obama faces would spell "marginal favorite" or "slight underdog," if -- and it's a big if -- the economy chugs along for the next few months at more or less the pace that most economists expect. Or, if Larry Bartels is right and it's a large plus if a president is sworn in when the economy really stinks, then Obama is a marginal-to-solid favorite (again, if the economy does as expected now).

Anyone who starts talking about "conditions in US that would normally..." really should know this stuff.

Now, granted, there's a big caveat in order. No one has ever run for re-election in exactly the same circumstances. No one else ever took office during the depths of a horrible recession, which then abated rapidly, but then never kicked in to a normal, strong, recovery (that is, with quarters of 5% growth or more). There simply is no direct precedent. What we're doing is extrapolating from sort-of similar situations as best we can, and prediction system skeptics are quite right to point that out. Still, on balance, I'd say it's a lot more wrong to call the current situation "conditions in US that would normally spell defeat for any incumbent" than it is to say that we have conditions that would normally make a president a marginal favorite.

(via Kraushaar)