An interesting take

Powerline speculates that Republicans have no Manhattan media presence except during election years, thus explaining their drops (and rises) in polls of the public, er, mind.

Sounds a little too convenient. I would have preferred seeing the observation from the conservative Powerline when Bush was riding high, rather than as an explanation for his riding low.

But it does strike me as an interesting, testable hypothesis. Is there a correlation between Republican approval ratings and elections? It seems obvious that at the presidential level, the answer is yes. Presumably many factors play into this kabuki, but the idea that one player remains relatively silent (involuntarily so, apparently) except during the relevant period is an interesting one.

Your mathematical model of public decision-making would have to have a reasonably short capacitance. But that seems likely enough, if perhaps somewhat variable. I suspect, for example, that "dishonest" is a necklace Clinton will wear forevermore in the public mind, while "Bush lied" has a shelf life of only five minutes after Brian Williams, Jimmy Carter or the NYTimes says it.