Adventures in Bad Poll Reporting

Usually Gallup does a very good job reporting on their own polls. This one, however, which shows that 53% of Americans blame Barack Obama either a great deal or a moderate amount for the nation's economic problems, is an exception. Their headline is "In U.S., Slight Majority Now Blame Obama for U.S. Economy." Huffington Post's headline was "Gallup: Majority Of Americans Blame Obama For Bad Economy," quite logical given Gallup's headline. But when I see that headline, I intuitively assume that it's one in the continuing series of polls asking who do you blame more for the nation's problems, Obama or the Republicans? It is not. Indeed, we can tell it is not because 69% blame George W. Bush a great deal or a moderate amount. I've seen polls asking Obama vs. (Bush and the Republicans). Gallup doesn't include current Republicans. The amount that I blame George W. Bush for the nation's current problems is going down, ever so slowly, with the amount that I blame John Boehner and his cohort going up much more quickly.

Also, of course, there's the standard complaint against this poll: the last three results on the same question were 50%, 48%, and 50% covering the last year-and-a-half. 53% is not a big change from that. It's not a big enough majority that we can be statistically confident that it is in fact a majority, or that there was any change from prior editions. But mainly my problem is that the headlines suggest that a majority place primary blame with Obama, when the poll continues to suggest the exact opposite.