I have a post at 538 updating my earlier analysis of Obama’s approval rating. Commenters both there and here, as well as Jay Cost and a couple others via email, noted that something seemed odd or amiss about the predictions I presented. As it turns out, they were right and in my post I discuss why.
The upshot of the new analysis: the discrepancy between Obama’s actual and expected approval rating is smaller than I originally reported (2.7 points vs. 9 points). However, in several recent quarters, the discrepancy between expected and actual approval has grown larger. It was 6 points in the first quarter of 2012.
The new analysis also goes further, showing how any discrepancy reflects greater-than-expected support among Democrats and independents, but not Republicans. I conclude:
bq. Of course, these results aren’t necessarily surprising. As the political scientist Gary Jacobson has documented, presidential approval ratings are increasingly polarized by party. This has certainly been true of Mr. Obama. But the very forces of partisan polarization — the “red America” and “blue America” — that Mr. Obama deplored as a presidential candidate may have buoyed his approval rating as president.
Of course, it’s a bummer to have to revise analysis after it was “published” — especially when it got some attention, e.g., via Memeorandum — but this is also a nice illustration of how crowd-sourcing can help you get things right. And given the tight timeline for the 2012 campaign book of which this analysis will be a part, crowd-sourcing may be a necessary form of peer review.
I thank commenters for their feedback.