The emergence of kinship behavior in structured populations of unrelated individuals

The paper provides an explanation for altruistic behavior based on the matching and learning technology in the population. In a infinite structured population, in which individuals meet and interact with their neighbors, individuals learn by imitating their more successful neighbors. We ask which strategies are robust against invasion of mutants: A strategy is unbeatable if when all play it and a finite group of identical mutants enters then the learning process eliminates the mutants with probability 1. We find that such an unbeatable strategy is necessarily one in which each individual behaves as if he is related to his neighbors and takes into account their welfare as well as his. The degree to which he cares depends on the radii of his neighborhoods.

Authors and Affiliations

1. Department of Statistics, School of Mathematical Sciences, Tel Aviv University, Tel Aviv, Israel, , , , , , IL

   Ilan Eshel

2. Department of Mathematics and its Applications, University of Naples, I-80138 Naples, Italy, , , , , , IT

   Emilia Sansone

3. Economics Department, Bonn University, 24 Adenauerallee, D-53113 Bonn, Germany (e-mail: shaked@econ3.uni-bonn.de), , , , , , DE

   Avner Shaked

Received June 1996/Revised version October 1998

Reprints and permissions