You said it. Especially since this is also on the gallup site ...
---
Bush campaign chief strategist Matthew Dowd's spin on our latest Bush job approval number -- 51% according to an Aug. 9-11 poll -- raised a few eyebrows among the sticklers for accuracy around this shop. Dowd reportedly said, "no president who has been at or above 50% at this point in an election year has lost."
Looking at historical job approval ratings from August of election years involving incumbents seeking re-election, the fact is, no president who has been at or above 52% at this stage has lost. But even that figure is misleadingly low because it is based on an outlier low point for Bill Clinton in 1996. For most of that period, Clinton's approval scores were in the 57% to 60% range. More broadly, no president who has had an average approval score of 55% or higher in the third quarter of an election year (spanning August to October) has lost.
Losing incumbents have had average approval ratings in the second and third quarters of an election year ranging from 35% (Carter and Bush I) to 46% (Ford).
Thus, while a 50% job approval would logically seem to be an important threshold for re-election, we just don't have a precedent to say anything definitive about whether that is the magic number. It could be 50%, but it could be as high as 55% or even as low as 47%.
Although Bush's latest rating of 51% is the highest he has achieved since April, it is within the statistical margin of error of the 47% to 49% approval scores he has consistently received since May. His average approval for the second quarter of 2004 (spanning May to July) was only 48%. In order to be in the company of successful incumbents, he will have to push that to 55% in the current quarter. If he continues to hover around 50%, he will be in uncharted territory, and the best one might say is that the election still figures to be extremely close.
