Actually the placebo effect is a statistical term covering the entire improvement seen in the placebo branch of a trial. Part of the effect comes from beliefs, yes. But there are other causes. For instance, people tend to enroll in clinical trials when their health is at a local minimum, and reversion to the mean can account for a good chunk of their improvement.