Last week, I predicted that President Obama's primetime speech to Congress would fail to have a significant effect on public opinion. While it's too early to reach a definitive conclusion, the early indications are largely consistent with that conclusion. An ABC News/Washington Post poll conducted Sept. 10-12 shows no statistically significant change in Obama's approval on health care or support for health care reform compared with a poll conducted August 10-12 (see also the Post story on the poll -- via Kaus). At best, Obama might have regained the ground he lost in late August -- a CBS News poll (PDF) conducted Sept. 10 showed a 12 point increase in approval of the president on health care compared with a poll conducted August 27-31, but that poll also showed no change in the percentage of Americans who think health care reform would help them personally. (In addition, the CBS poll re-surveyed respondents from the August 27-31 poll -- a format that is useful for comparing opinions before and after the speech, but may not be fully representative.)
Update 9/16 8:36 PM: Nate Silver calls me and George Stephanopolous out, falsely stating that both of us "[concluded] that there is no bounce on the basis of the ABC poll... while ignoring the other polling." That's wrong on two counts. First, at the time I posted, I had not seen any post-speech polls other than the ones cited in the post. Second, I didn't say "there is no bounce" -- I said the speech would most likely "fail to have a significant effect on public opinion" and that, "While it's too early to reach a definitive conclusion, the early indications are largely consistent with that conclusion." (Note also the post title: "Obama's health numbers: Not moving much" [emphasis added].)
Since I wrote that post, Rasmussen and CNN have released polls showing what Mark Blumenthal describes as "[s]mall, nominal increases in approval for Obama or support for health reform." In particular, the observed increases in support for health reform in the two polls were not statistically significant (Rasmussen's has seemingly dissipated already). Despite preliminary evidence of a small uptick in Obama's approval, I'll stand by my claim.