Knowledge sharing and rewards: a game-theoretical perspective

Abstract

Shared databases are used as tools for knowledge management. But contributing knowledge to the database is a public-goods dilemma. Its specific features are multidimensionality of the goods, resource overlap, interdependency of group members and the lack of a marginal unit of return for individual contributions. An analytical model is presented describing the individual- and group-level payoff function. It shows that rewarding contributions with a cost-compensating bonus can be an effective solution not only at the individual level, but also at the group level. Whether or not a bonus system is self-supporting depends on the critical mass of people on average needing a contribution. This depends on the overlap of the people's resources and their interdependency.

Keywords:

• knowledge sharing
• knowledge exchange
• knowledge utilization
• social dilemma
• knowledge management strategy
• self-supporting bonus system

Notes

1 In this article, we are mainly interested in a formal analysis of the dilemma. We do not distinguish between knowledge and information, even though we are aware that from a cognitive point of view there are some distinctions.

2 The Prisoner's Dilemma first was described and formalized by Albert Tucker (1950). It describes a situation where two people are arrested by the police. The police have insufficient evidence for a conviction and offer each of the suspects the following deal: if he testifies for the prosecution against the other and the other remains silent, he goes free and the silent accomplice receives 10-year sentence. If both stay silent, the police can sentence both prisoners to only 6 months in jail for a minor charge. If each betrays the other, each will receive a 2-year sentence. The dilemma results from the fact that for each person it is always better to testify: if the other testifies he receives 2 years instead of 10 years and if the other stays silent he goes free instead of 6 months. But if both chose this – individually seen – best way and testify than they both receive 2 years, which is much more than if both cooperate and keep silent.

Additional information

Notes on contributors

Ulrike Cress

About the authors

Ulrike Cress received her University degree in Psychology 1994 at the University of Tuebingen, Germany, and her Ph.D. in 1999. Since then she is assistant Professor at the University Tuebingen, and senior researcher at the Knowledge Media Research Center (KMRC). She wrote papers about knowledge exchange, social dilemma, e-learning and evaluation.

Stefan Martin

Stefan Martin received his University degree in Mathematics at the University of Wuppertal, Germany. Since 2002 he is a research associate at the Knowledge Media Research Center (KMRC) in Tuebingen. His research interests are mathematical modeling of communication processes and simulations.