Generalized exchange and social dilemmas

Citation: Toshio Yamagishi, Karen S. Cook (1993) Generalized exchange and social dilemmas. Social Psychology Quarterly (RSS)
Internet Archive Scholar (search for fulltext): Generalized exchange and social dilemmas
Tagged: Sociology (RSS)

Summary

Acawiki Summary

Toshio Yamagishi is famous for experimental examinations of social dilemmas (e.g., having people play game theory games). Karen Cook is most well known for the central role she has played in the development of exchange theory during the eighties. As the title implies, the primary contribution of this paper is to look at the interactions between the two fields or areas.

The paper begins with a description of generalized exchange. While described in details first by Ekeh's (1974) Social exchange theory: The two traditions, this paper builds up a detailed description of generalized exchange so that it can be tested experimentally. The paper introduces two structures of generalized exchange. One involves a commons resource pool where every contributor can give some resource and where resources will then be magnified (e.g., through investment) and then given back evenly to each of the participants. The second model is a network based structure where each participant gives to one other participant who gives to another participant in a sort of daisy-chain arrangement and where the resources given is similarly magnified. The authors describe the way that these structures can resemble n-person prison dilemmas (the first structure) or assurance games (the second structure) games or prisoners dilemma.

Building on Cook's work on trust, the authors also incorporate a measure of how likely players are to trust others.

The authors propose a series of hypotheses (each quoted verbatim below) that aim to explore the relationship between structure and actions in social dilemmas:

• H1: Participants will cooperate more in the network-generalized exchange structure than in the group-generalized exchange structure.
• H2: High trusters will cooperate more with low trusters in both structures.
• H3: The positive effect of trust predicted by Hypothesis 2 will be stronger in the network-generalized exchange structure than in the group-generalized exchange structure.
• H4: Information about the other participants actions will have a positive effect on levels of cooperation in the network-generalized exchange structure, but this positive effect of information feedback is not expected to emerge in the group-generalized exchange structure.

The results used four-person groups playing a PD and AG type game within the network structures described in an experimental setting. The results showed strong support for the first two hypotheses, and support for the third but no support for the fourth.

The authors suggested that large groups with partial information might be more susceptible to breaches of trust and might give us less in the network structure. In fact, the authors found that the opposite was true.

Theoretical and Practical Relevance

The paper has been cited more than 140 times in the first 17 years since it's publications and continues to be one of the core citations in the sociological and social psychological literature on generalized exchange.