The power of requests in a redistribution game: an experimental study

Abstract

In most situations of voluntary contribution people are willing to give at the beginning, however contribution rates decay over time. In a new setup we introduce non-enforceable sharing rules, as requests, in a repeated redistribution game (called tip pooling). Three experimental treatments differ by the requested amount of sharing of privately known random endowments (tips), with one player never receiving any endowment. Using a hurdle model, we find no significant difference in free riding between the three sharing rules, but strong differences in positive contributions which, however, are lower than the rules prescribe. Furthermore, the average positive contribution remains stable over time, while free riding tends to increase.