Spinning. And The Washington Post just can't quit quoting false claims without noting the facts or naming the claimer.
In an article about how Clinton aids are (anonymously) admitting that Clinton can't win the nomination, Perry Bacon Jr. and Anne E. Kornblut quote a vague source in a way which makes a false claim their own.
The Clinton campaign has tried to sway voters and superdelegates for weeks by pointing to opinion polls that show Obama's favorability ratings steadily decreasing since his string of victories in February.
According to Pollingreport.com the maximum across pollsters of the difference between Obama's highest favorable rating and his latest favorable rating is 8% (the minimum is zero). Across cross pollsters of Obama's average favorable rating in April and May minus that in February (his best month overall) is CBS -3.5 % USA Today/Gallup -0.5 %, CNN -2%, Newsweek no poll in Feb from March (best month) -6%, ABC Wapo no poll in Feb down 7% since Janauary others none in April/May and or none in Feb with Obama at his all time peak.
Notice that unless I change my rule in course (for Ballance as the Post does) I get only tiny declines.
The CBS/New York Times poll ends with an uptick from April to May making the claim of a steady decline technically false. Of course it is undefined and Bacon, Kornblut and their vague sources can argue that the decline would be steady if it were smoothed enough, which is, you, know tautological as it just means that although it isn't steady, it can be steadied, that is, although it isn't smooth it can be smoothed. The word "steady" was chosen to be technically true but misleading as it was meant to suggest "strong" or "enough to make one pessimistic about his chances in November". However, if one is aiming to be technically true, one has to be precise (adding "averaged across pollsters" or "for all but one pollster").
Now I think Bacon and Kornblutt would argue that the main thrust of their article is that Clinton staffers are admitting that it is over, so a little distortion pointing in the opposite direction does no harm. I would ask, why not the truth. The claim is technically false and deliberately misleading. It should not be quoted without attribution or contestation.
update: links corrected thanks to Brendan