An experimental study on the information structure in teams

Is free-riding in teams reduced when one member receives a signal on his colleagueís performance? And how does free-riding depend on the signal's type? We address these questions in experimental teams in which two agents sequentially exert effort to contribute to the team output. We vary the type of information the second mover receives prior to his effort choice and find that agents work more when signals are available. Overall, behavior differs from predictions of standard theory. Signals that are predicted to have no effect are, in fact, influential and signals that are predicted to have an effect are redundant.