I'm beginning to think Mr. Duncan might have a rather large ego himself.
Maybe this is a stupid question. But some of the statistics I've seen show new teachers (0-3 years) being fired at higher rates than experienced teachers.
But isn't it possible that the discrepancy here is because the lousy teachers never go on to be experienced teachers? How can we really have a valid measure of this? Anecdotes pack a powerful punch, but not much in the way of broad information about the system there.