Andrew Sullivan continues to freak out:
Here's a dishhead bleg: when was the last time that a sitting president in a re-election campaign lost six percentage points in the polls in two weeks in October?
That's pretty specific, and I suppose the answer is "never." But let's change the question: When was the last time Barack Obama lost six percentage points to a Republican challenger?
Answer: 2008. On September 1st, Obama led John McCain by 6.4 points. On September 10th, McCain led Obama by 2.5 points. That's a swing of 8.9 points. But when the election was actually held, Obama won by 5 points.
Look: Polls change. That's politics. Beyond that, though, there are a couple of specific reasons everyone should settle down a bit. First, the fundamentals have always suggested that this would be a close election. The consensus of the political science models is an Obama win by maybe 2 points or so. Second, I'm increasingly convinced that a couple of years from now some enterprising political scientist will write a paper thoroughly debunking the idea that Obama's debate performance was as horrible as everyone is making it out to be. Instead, the recent poll changes will come down to three things:
- A late September surge by Romney for reasons that are (at the moment) still a bit of mystery.
- Reversion to the mean. Obama was never going to win the election by 5 or 6 points, and his recent drop has been baked into the cake for a long time. His big lead was mostly an artifact of stupid mistakes by Romney, and eventually Romney recovered from them.
- The media freakout over Obama's debate demeanor.
Obama didn't turn in a great debate performance, but it was nowhere near bad enough to account for the kind of poll declines we've seen over the past couple of weeks. That's my two cents, anyway.