Thursday, August 27, 2009


MR linked to this fascinating survey piece on recent research on the placebo effect from Wired magazine.

Back in graduate school, Lester Telser circulated an article in his class of a randomized control trial of four treatments: branded aspirin, unbranded aspirin, branded placebo and unbranded placebo. The effects were all distinguishable and in the order listed. This has really important implications for welfare analyses of advertising (among other things) which I do not think have yet been incorporated into the literature.

It is interesting to think about placebo effects in economics. Given the important role of expectations in thinking about business cycles, one might think about, say, fiscal stimulii applied early on in the downward part of the business cycle as being aimed in part at producing placebo effects operating to change the mood, and thus the behavior, of consumers and investors. It is a bit tricky to distinguish between changing expectations about the actual likely path of future outcomes from simply changing the mood of agents in the economy. Suppose, for a moment, that fiscal stimulii actually have no real effects but that all consumers and investors believe that they do. In such a world, does a stimulus have a real effect or a placebo effect or both?

Another interesting question concerns the welfare effects of bans on patent medicines that are not directly harmful but not directly effective either.