Mark Schmitt describes the willingness of people to place their faith in President Bush in terms of cognitive dissonance without describing it as such:

But there’s a factor here that I think is always neglected in these polls: There’s always a thin line between what we believe and what we wish to be true, or need to believe to be true. Unless you’ve given up completely on the Bush presidency, you naturally want to think the president and the government are doing the best job they can do with the disaster, because to believe otherwise is kind of terrifying.

Like I always say, cognitive dissonance makes the world go round. In this case, the two cognitions in conflict are that we’re totally screwed if the President is really bad at his job and indisputable evidence that the President is really bad at his job. It’s easier for many people to ignore the evidence than to believe that we’re totally screwed.