Does conditional cooperation in public good games depend on the marginal per capita return

What is it about?

In this paper we explore conditional cooperation in public good game games. The standard way of measuring conditional cooperation was introduced by Fischbacher, Gächter and Fehr (2001, Economics Letters). It involves asking the subject to fill in a contribution table in which they detail how much they would contribute to a public good as a function of the average contribution of other members of the group. A subject is labelled a conditional cooperator if their contribution is positively correlated with the average contribution of others. Typically, around 50-70% of subjects are conditional cooperators. Conditional cooperation has conventionally been measured when the marginal per captia return (MPCR) to the public good is 0.4; i.e. every dollar contributed to the public good yields 0.40 for every member of the group. The main objective of our paper is to see how conditional cooperation changes if the MPCR varies. Specifically we compare an MPCR of 0.4 with 0.8 and 0.2. Note that with the low MPCR of 0.2 it is not efficient to contribute to the public good. We find that rates of conditional cooperation remain relatively stable for different values of the MPCR. We also find, though, that unconditional contributions are significantly influenced by the MPCR. Specifically, subjects giving very little when the MPCR is 0.2 and a lot when it is 0.8. Note that this ultimately means conditional cooperators will end up contributing a lot less when the MPCR is 0.2 than 0.8. Our findings challenge to some extent common interpretations of conditional cooperation and suggest that cooperation is driven more by imitation and leadership rather than social preferences. A second objective of our paper was to see if conditional cooperation is influenced by the strategic timing of the game. In the standard way of measuring conditional cooperation a subject conditions their contribution on that of three other people; in interpretation it is as if the subject chooses after everyone else. We compare that with a situation where a subject conditions their contribution on that of a leader. In this case the subject only sees what one other member of the group has done. We find that conditional cooperation, and unconditional cooperation, are not influenced by this change of game. This is a reassuring finding given that results on conditional cooperation have been applied to look at leadership.

Featured Image