Agent-based analysis of agricultural policies: An illustration of the agricultural policy simulator AgriPoliS, its adaptation and behavior

This paper combines agent-based modeling of structural change with agricultural policy analysis. Using the agent-based model AgriPoliS, we investigate the impact of a regime switch in agricultural policy on structural change under various framework conditions. Instead of first doing a sensitivity analysis to analyze the properties of our model and then examining the introduced policy in an isolated manner, we use a meta-modeling approach in combination with the statistical technique of Design of Experiments to systematically analyze the relationship between policy change and model assumptions regarding key determinants of structural change such as interest rates, managerial abilities, and technical change. As a result, we observe that the effects of policies are quite sensitive to the mentioned properties. We conclude that an isolated analysis of a policy regime switch would be of only minor value for policy advice given the ability of simulation models to examine various potential futures.