I agree with Tyler that there is some serious evidence for placebo effects, especially although not exclusively for subjective components of disease. But the evidence is usually overstated because it is confused with the natural tendency of sick people to get better. A typical medical study, for example, will compare the results of a new drug against a placebo. The improvement in health of those on the placebo is then labeled “the placebo effect” – but this is wrong. To correctly identify the effect of the placebo one needs three randomly selected groups – a treated group, a placebo group and a non-treated group. The effect of the placebo per se is then measured by the health differences between the placebo and non-treated group. Although spontaneous healing effects are large, placebo effects when measured correctly tend to be small although not non-existent.