To examine the claim that rewards are effective at altering behavior, we pose three questions: First, for whom are they effective? Second, for how long are they effective? And third, at what, exactly, are they effective? (I have already hinted at a fourth question--At what cost are they effective?--but