Agent-Based Simulation of Local Soy Value Chains in Ghana

Abstract

The assessment of changes in the relationships between supply chain agents is considered fundamental for market transformation. This paper reports on the application of a Value Chain Lab that supports the measurement of behavioral change in vertically structured supply-chain relationships. A participative gaming approach is used that enables to identify changes in mutual trust, transaction costs and risk behavior that result from value chain support and co-operation. The Value Chain Lab comprises value chain analysis, value chain games and multi-agent simulation. The paper describes the multi-agent simulation of a soy value chain in northern Ghana. The research was conducted in the context of the 2SCALE program, aiming to improve rural livelihoods and food and nutrition security in a number of African countries by developing agricultural supply chains including local smallholder farmers. The study confirms the positive effects of trust and loyalty in value chain relationships. Furthermore, it demonstrates the usefulness of agent-based simulations for exploring potential consequences of alternative interventions.