Revisiting Kindness and Confusion in Public Goods Experiments: Comment∗

We present a novel experiment to evaluate the role of confusion in explaining the dynamics of contributions in public goods experiments. In a learning condition we keep the subjects confused by withholding the information that would be necessary for them to understand the game. The comparison of the results with those of a standard public goods treatment provides a lower bound for the influence of confusion on contribution dynamics. We find that learning in the state of confusion explains 41 percent of the contribution dynamics in the standard public goods game. This result complements that of Houser and Kurzban (2002) who found that all of the decrease in contributions can be attributed to the reduction of confusion. We argue that their findings can be seen as an upper bound for the influence of confusion.