Project Details
Projekt
Impacts of strategic behavior on land market dynamics and regulation effects
Subject Area
Agricultural Economics, Agricultural Policy, Agricultural Sociology
Term
since 2017
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 317374551
Land market transactions are strategically relevant for farms. First, these transactions affect directly their future development perspectives because these depend on the farm’s current land bank well as on financial obligations linked to the transactions. Second, the future evolution of any farm is interdependent with the evolution of neighboring farms because their relative competitiveness and local competition for land affects growth, stagnation, shrinkage or exit. We will address both arguments through a) explicit modelling of strategic land rental behavior in an agent-based model, b) subsequent experiments on the differences in the performance of strategically acting farms and their neighbors, and c) evaluating the effects of various land market settings and political framework conditions.a) Modelling strategic land market behaviorAlthough agent-based modelling of the agricultural sector has become widespread, land market interactions are usually ignored or modelled in a simplified way based on ad hoc assumptions. For instance, within the agent-based model AgriPoliS, farms engage in the rental market according to a myopic calculation of shadow prices. To overcome the limitations of myopic behavior, we will apply deep neural networks (DNN) for modelling strategic behavior of farms within AgriPoliS. The DNN will replace the current land market, investment and exit decision making of one, several, or all agents. The DNN will process available information of the respective farms as well as information about neighborhood and market conditions. The DNN will be trained using reinforcement learning in repeated experiments for alternative settings in order to optimize long-term performance. b) Economic analysis of strategic land market behaviorWhether a farm exits, shrinks or expands via the land market depends on farm characteristics, political and market conditions, the local competitive situation as well as the decision process. Experiments comparing the outcomes of myopically optimizing versus strategically acting agents within various initializations of AgriPoliS will be used to assess the effects of strategic actions on the individual farm as well as on the local/regional level. Beyond information on performance, we will use these results to assess the economic sustainability of strategic market actions with regard to the farms’ resilience as well the emergence and nature of market power.c) Structural and policy implications of strategic land market behaviorStrategically acting and interacting farms may affect structural change via path breaking behavior in various forms as well as the exploitation of market power. Land market regulation may intervene in order to reduce unwanted effects such as strong concentration of land. We aim to study the effects of regulation scenarios for alternative case study regions.
DFG Programme
Research Units
