Purpose: To examine the effect of individual versus group evaluation and reward systems on work group behavior and performance under different task conditions. Methodology: Uses computational social methods using Agent Based Models to simulate work group interactions as different forms of iterated games. Findings: Group based systems outperform individual based and mixed systems, producing more cooperative behavior, the best performing groups and individuals in most types of interaction games. A new role emerges, the self-sacrificer, who plays a critical role in enabling other group members and the group, to perform better at their own expense. Research Implications: Suggest opportunities for model development and guidelines for designing real world experiments. Practical Implications: Helps firms engineer better performing work groups as well as the design of other business systems. Social Implications: Identifies mechanisms by which cooperation can be developed in social systems. Originality/Value: Demonstrates the role and value of computational social science methods and agent based models to business research.
- Agent based models
- Complex systems
- Computational social science
- Group versus individual reward systems
- Work groups