In a recent paper in Journal of Economic Behavior and Organization, Croson (2000) shows that when subjects' beliefs about the contributions of others are elicited in a voluntary contributions public goods game, those subjects contribute less than do other subjects in the same game when beliefs are not elicited. We report briefly on a similar experiment where we do not observe this. There are small differences between this experiment and Croson's that may explain these different results, but none of these explanations are completely satisfying. So while Croson shows that belief elicitation procedures may have unwanted effects, this does not always occur; and it may be difficult to predict when and where belief elicitation procedures will have such undesirable effects.
