An experimental inquiry into the nature of relational goods

Our experiment aims at studying the impact of two types of relational goods on the voluntary contributions to the production of a public good, i.e. acquaintance among the contributors and having performed a common work before the experiment. We implement two treatments with 128 participants from two different groups. In the first treatment the subjects are left talking in a room before the experiment (cheap talk treatment); they are not suggested any particular topic to talk about, nor are they requested to perform any activity in particular. The second treatment involves the performance of a common work (namely, the computation of some indices of economic performance of three companies, based on their balance sheets). The two groups of subjects are composed either by people with or without previous acquaintance. An equal number of subjects from each of these groups is then allocated to either treatment. After that the subjects played a standard 10-rounds public goods game in groups of 4. The groups were gender-homogeneous. This allows us also to inquire for the possible presence of a gender effect in our experiment. Our results show that: 1) both common work and previous acquaintance increase the average contribution to the public good, 2) there is a relevant gender effect with women contributing more or less than men, depending on the treatment. Therefore, we conclude that relational goods are important to enhance cooperation, that acquaintance and working together are rather complements than substitutes, and that different relational goods produce different effects on cooperation. Also, we find further evidence for women's behaviour to be more context-specific than men's.