Yesterday, Gallup reported the first real discernible signs of a convention "bounce" in Obama polling. Over the course of the week, Obama's job approval rating moved up 7 points. In Gallup's daily tracking Friday, 52% of adults approved of the job Obama was doing. In a head-to-head match-up, Obama led Mitt Romney by 3 points, 48-45. It is clear that up to that point in the convention, Obama was enjoying a bounce. It also seems clear, however, that his bounce will deflate rather quickly.
Gallup's daily tracking of job approval is based on a rolling three-day average. Friday's number, then, is based on interviews conducted Tuesday, Wednesday and Thursday. Obama didn't speak until well after 10pm EDT, so most of the interviews on Thursday would have been conducted after he spoke. As a result, Friday's number is based largely on feedback to speeches by First Lady Michelle Obama and former President Bill Clinton. These speeches were the high-point of the DNC's convention. While Clinton's speech was largely a work of fiction, his performance of it was outstanding.
Given two solid speeches and the media's loud howls of approval one would expect a bump in the polls for Obama. What Friday's number doesn't reflect, however, are Obama's lackluster and pedestrian performance in his acceptance speech and the dismal Friday jobs numbers.
How bad was Obama's speech? Just 15 minutes after he finished, Rachel Maddow and her crew spent all their time talking about Mitt Romney and Afghanistion. Virtually every media pundit was forced to admit that, at best, it wasn't one of Obama's best speeches.
Combine that with the very bad jobs report on Friday and we should expect to see Obama's job approval number come down very rapidly next week.
Follow me on twitter
PS: I can't fail to stress again that Gallup's job approval number is based on interviews with adults. While this might be of fleeting interest to a sociologist, it really doesn't tell us much about the upcoming election. Polls of adults skew heavily in favor of Democrats. Beyond that, it isn't particularly useful to put too much weight on a number based on a not-insignificant number of people who won't be voting. Looking at long-term trends in this number can, however, be interesting.