Frank Newport the Gallup Guru of USA Today has an interesting article in which he identifies a hint of positive news regarding George W Bush's poll ratings. They remain, of course, the worst since the Carter administration. But Newport argues that they are on the rise for the first time in a long time. The last two Gallup polls have shown the President's approve/disapprove ratio improve. Until July they had slipped for months.
That 34% of Americans approve of the way the President is doing his job can only be seen as good news when compared with the low of 29% in early July. He is now back where he was in the early part of the year, just after the Democrats had taken over Congress.
With a three point margin of error this rise of five points could be a fluke. The 29% recorded a month ago might really have been as high as 32% and the 34% of the latest poll might have been as low as 31%. That would mean no rise at all. But Gallup has recorded a rise in two successive polls, so I thought it wise to explore further.
I looked at the President's poll ratings in surveys produced by nine organisations: Gallup; Rasmussen; Newsweek; NBC/WSJ; Pew Research; CBS/NYTimes; Hotline/FD; ABC/Washington Post; and Fox News. In each case I examined the last four polls, though that was not perhaps the most scientific method, as the gaps between the polls are not equal. Gallup, which recorded a fall from the first to the second and then rises from second to third and third to fourth was not atypical.
In fact, of the nine polls, only one showed the President's approval rating declining in its most recent poll: and that was ABC/Washington Post, which has not published a poll since mid July. In other words, the ABC/WaPo poll shows a nadir that is only just after the nadir shown by Gallup, but Gallup has shown two rises since its recorded nadir.
Two polling organisations showed stability between their third and fourth polls. Six showed the President's approval rating was up.
In no case is the current ratio significantly better (and it is often worse) than in the first of the four poll series that I examined. But this too bears out Newport's suggestion that there was a nadir in the first half of July from which the President has since recovered.
Is the improvement real? And is it the start of a trend? I think the answers are "yes" and "it's too early to say". In particular the fact that six out of nine polls have shown an upturn and only one not very recent poll a downturn is persuasive. But a trend that is only three weeks old is hardly a trend at all.