We create an agent based model on farm growth. Farmers who work on homogeneous landscape are rational decision makers and try to expand their farms by taking over neighbouring farmland. Through selling and buying farmland between farmers, the model captures the number of farmers and average farm size over time. These two indicators are then used to analyse the change of the land use system. The underlying driver is that due to increasing agricultural productivity and declining prices farmers need more and more land to economically maintain their farms. Farmers who are unable to grow and who do not have a successor will eventually quit. Sub-models include succession, make-a-transaction, sell-to-expanding-neighbours, and are coordinated by a decision tree. Our goal is to recreate the s-shaped growth of U.S. farms over the 20th century, which according to classical explanations was driven by economic growth, technology development, and market integration, while we argue that interactions and feedback mechanisms may also play an important role.
Current work is on the psychological model of decision making processes of farmers which integrates interactions and feedbacks. We reject farmers being rational decision makers but involving in social interactions to make land use decisions. We define the scale of interaction, types of interaction (strategies) and possible feedbacks. The model will be used to provide explanations of regime shifts in land use systems (from another PhD’s contribution on empirical studies), if validation is allowed.