My friend Scoot sent me a note on a WaPo article by the decidedly left of center E. J. Dionne. Now I generally do not like Dionne, I think his analysis tends to be shallow and his insight weak. I think this article is no exception. Within the OpEd pieces he details the woeful state of polls and opinions about the present administration. He quotes a number of political groupies and a congressman and proclaims Armageddon for right wing views and republicans. Bets are for suckers. Not that one shouldn’t have an idea based upon some testable facts, but most social circumstances suffer from nearly limitless variability. This includes weather, sports scores and the phase of the moon. Check out Freakonomics for more on these effects.
Polls are (in my opinion) next to useless. They present a very lopsided view view of most issues (in that they begin as biased instruments.) I prefer to draw conclusions from observed behavior. In an earlier time, people claimed to like Betamax better than VHS, but they bought VHS. The power of the marketplace is more telling than any focus group Sony ever conducted.
I would submit that The President's approval ratings are largely inconsequential. As of now, he is effectively out of office and thus has nothing to lose. As his VP will not be running in 2008, the administration essentially has a pass on popularity. Part of this means that the expectations of the public (I think) are higher. If you have nothing to lose, then you should be going whole hog, right?
I think the entire fixation over national polls around the President will have little to no effect on the midterm elections. Folks are very locally focused. I would predict a 99% success rate for incumbents. Despite the unhappiness with the present set of policies (which in my opinion are not actually bad and in many ways good), most partisans will fight very hard locally to retain "their guy" in Congress.
And except for super safe seats, they simply won't ask President Bush to help them.