Tuesday, 6 December 2005

What a gas

The president’s poll numbers appear to be recovering as of late, and there are two major competing theories to explain the change. Charles Franklin appears to attribute the change to the new PR pushback from the White House, which we might term the Feaver-Gelpi thesis (see also Sunday’s NYT), while Glenn Reynolds says it’s the gas prices and the Mystery Pollster suggests good economic news in general.

It may be the most simplistic thesis, but I think the “pump price” explanation is probably the most plausible; unlike other information, gasoline prices are unavoidable information for most voters and not subject to partisan spin, unlike the presidential pushback on Iraq and news of the general economic recovery—both of which can be spun negatively in a way that falling gasoline prices really can’t. In a noise-filled informational environment, I suspect clear “pocketbook” signals like gasoline prices are much stronger cues for presidential support than the world of competing, ideologically-based claims over Iraq and interest rates.

Update: Al Qaeda appears to put some stock in the pump price explanation as well.

1 comment:

Any views expressed in these comments are solely those of their authors; they do not reflect the views of the authors of Signifying Nothing, unless attributed to one of us.
[Permalink] 1. Scott wrote @ Wed, 7 Dec 2005, 8:07 am CST:

I think the pump price may be the driving force, but I believe the overall decline was due to an interactive effect from multiple sources that, as a whole, were greater than the sum of their parts…factors which…individually…may not have been enough to cause a serious slide.

I think the fallout from Katrina (other than gas prices), continued SERIOUSLY bad news in Iraq, and gas prices combined to cause a drop steeper than gas prices or (non-gas) Katrina fallout alone would have caused. I also think it is possible that bad news in Iraq would not have caused a significant drop (or possibly any) without the effect interacting with impressions of government failure in LA and MS. This, in my opinion, would have a doubly negative effect….(1) the effect would be stronger among the most sophisticated and attentive (controlling for PID) and (2) the effect would actually SHOW UP in the non-sophisticates…whereas any of the three on their own would have had little or no impact among this group.

The first point fits nicely with the, dare I say, brilliant “update” model of evaluation and the second, if correct, hints at the existence of some type of impact threshold that needs to be crossed for converging events to significantly impact the opinion of non-sophisticates. I can’t recall, without checking a couple of references, if anyone has done any work on this.

Comments are now closed on this post.