[T]he administration is not politically deaf. Bush and his advisers can hear the rumblings of concern in the public and within their party's own ranks, and last week they began taking steps to shore up support for the war. In the view of the White House, the public is periodically upset by the violent images on its TVs and so the president must, from time to time, speak up. The model for the president's speech this week was his address to the U.S. Army War College in Carlisle, Pa., last year. . . .
[B]ut it's not clear that Bush's speeches serve to inspire. According to the Gallup poll, support for the war in Iraq went up 1 percentage point after his War College speech last year. Public confidence seems to more closely track the ebbs and flows in violence.
First, a 1 percentage point change is within the margin of error of the poll -- it's literally meaningless. But the larger point that Newsweek is driving at is clear. In the aggregate, presidential approval responds to fundamentals -- the economy, the number of war deaths, prospects for victory, etc. Prime time speeches just don't work. As George Edwards suggests in his book On Deaf Ears, presidential initiatives to change public opinion rarely succeed. What's ironic is that every president has to re-learn this lesson. Bush's particular lesson has been delayed for years by the popularity boost he received from 9/11, but he will soon realize that his top second-term domestic and foreign policy initiatives -- Social Security and Iraq -- are essentially immune to his efforts to whip up support.
In short, the perception that public opinion fluctuates wildly in response to elite messages is wrong, at least when elites are divided. Macro public opinion is actually quite stable, and it moves in understandable long-term patterns. So go democracy! In the aggregate, the public doesn't listen to spin; they respond to fundamentals.
(For more on the specifics of public opinion toward war, see Choosing Your Battles by my colleagues at Duke, Peter Feaver and Chris Gelpi. The bottom line of their research is that the public is willing to tolerate a relatively high number of casualties if they think victory is likely and the cause is important. And for more on macro public opinion, see Page and Shapiro's The Rational Public and especially Erickson, MacKuen, and Stimson's classic The Macro Polity.)