Pushed by the crowd or pulled by the leaders? Peer effects in Pay-What-You-Want

Literature on charitable giving often finds that seed money matters: the example of a wealthy donor is followed by others (List and Lucking-Riley, 2002). Nearly all relevant theoretical accounts (e.g. that leaders possess superior information on quality of the project) seem to apply to the closely related environment of Pay-What-You-Want mechanisms as well. Yet, as far as we can tell, no empirical study has tested for that until now. To fill this gap, we analyze data from 16 campaigns of BookRage (an equivalent of Humble Bundle, offering bundles of e-books). We make use of the fact that a fixed number of currently highest contributions are always displayed (along with mean contribution and total amount raised). Thus a discontinuity may be expected: contributions that are displayed might directly affect subsequent donors' behavior, in contrast to just slightly lower donations that are only observable as a (small) change in mean contribution. We find that the example of leaders makes no impact on willingness to purchase and amount paid. By contrast, the mean of past contributions has a positive impact on current contribution, yet a negative impact on the probability of contributing.