Heavy Lifting (ht: Jim Miller on Politics) brings us the graph above which charts Bush's approval ratings with the inverse of gasoline prices (the original source is Professor Pollkatz) and comments:
One of my current econometrics students is working on a similar hypothesis, trying to determine which is more important to presidential approval, accumulated U.S. deaths in Iraq or the price of gasoline. While the talking heads and the far left would have us believe it is the former, it is all too likely that it is really the latter.
Is the price of gasoline really the most important criteria by which the American public decides whether or not we approve of our Presidents? Are we that shallow?