In this paper, we use unique data from the market for Bordeaux wine to test the hypothesis that consumers are willing to pay for expert opinion because it is accurate. Using proprietary indicators of the quality of the vintage, which are based on both publicly and privately available information, we find that additional publicly available information on the weather improves the expert's predictions of subsequent prices. This establishes that the expert opinions are not efficient, in the sense that they can be easily improved, and that these opinions must be demanded, at least in part, for some purpose other than their accuracy. (JEL Classification: D8, Q13)
