A model of sequential effects in common pool resource dilemmas

Experimental games are often used as models of social dilemmas in which small groups of players have unrestricted access to, and share, a common and finite resource of desirable goods in the absence of any binding coordination mechanism. Examples are fishing, hunting, and use of computational facilities. Experimental studies of social dilemmas employ different protocols of play that differ from each other in terms of the information available to the players when they register their requests from the common resource. In this study we focus on the sequential protocol, where each participant has complete information about his or her position and the total requests of the previous movers, and the positional protocol, where each player only knows his or her position, but has no information about the other's requests. Previous research has found a robust position effect: individual requests are inversely related to the players' positions in the sequence with the first mover requesting most, and the last mover requesting the least. In an attempt to characterize the nature and intensity of the position effect, we developed and tested a descriptive model with one free (individual specific) parameter. The parameter is estimated from the players' requests under the positional protocol (i.e. in the absence of any information about the other's requests) and, as such, quantifies the social norm of 'advantage of the early mover'. In a range of cases examined, including different group sizes, resource pool sizes, and positions, the model predicted very accurately individuals' requests in the sequential protocol (with full information about position and others' requests). Copyright (C) 2002 John Wiley Sons, Ltd.