2019 Volume 8 Issue 2 Pages 19-33
In this paper, we consider indirect reciprocity which is an important mechanism on the evolution of cooperation asking why the humans can voluntarily cooperate with others in social dilemma situations. There is a severe conflict between theoretical studies and empirical ones in the aspect of what information is required for assessing others. Theoretical studies claim that complex information should be required because an assessment rule using behavioral information only cannot have an evolutionary stability while empirical studies object because many experiments show that they may not proceed such the complex information. Here we analyzed a private assessment system relaxing a public assessment assumption which is unrealistic but often used in theoretical analyses. Its rigorous analysis is extremely difficult because the private assessment system must solve a system with infinite equations. To do so, we introduced another assumption to solve it strictly. Moreover, we complementarily performed an agent-based simulation to confirm a reliability of the solution due to test the effects on the solution of the assumption introduced. As a result, we show that an adaptive norm on indirect reciprocity to keep cooperative regimes in a private assessment system has different features from well-known norms considering in a public assessment system. We also show that the staying norm we defined for resolving the issue actualized in the private assessment system has a superiority. To confirm our result empirically, we conducted subjects experiments and tested allowance of the staying norm statistically. Integrating theory, experiments, and simulations to explore an adaptive norm on indirect reciprocity may be possible to provide a new contribution to computational social science.
