Simulation of smallholder farmers’ adaptation to irrigation-induced landslides: An ABM application in human-land-policy interaction

Abstract

Understanding the dynamic interaction between smallholder farmers' adaptation decision-making behavior and the policy environment in the context of human-induced disasters is vital for developing effective policies to enhance farmers' participation in disaster risk reduction (DRR). From the perspective of bounded rationality theory, this study constructs the farmer-land-policy interaction (FLPI) model, taking Heifangtai in Loess Plateau, China, as a case study. Based on first-hand micro data, this study combines static econometric model analysis with dynamic simulation techniques, achieving a combination of micro decision-making and macro policies to understand the interaction between farmers and the policy environment. The FLPI model simulates farmers' future adaptations under multiple policy scenarios (baseline, mandatory policy, and subsidy policy scenario with a combination of different policy tools) from 2021 to 2030. Compared with the simulation results of the baseline scenario, the results indicate that mandatory and subsidy policies are effective interventions incentivizing farmers’ adaptation behavior for effective DRR. Mandatory policies rely on policy intervention to ensure that farmers adopt drip irrigation and to promote the generation of long-term behavior of farmers. And the results also propose a suitable subsidy range of 700–900 yuan and with a period of more than 2 years for subsidy policy. This study enriches the research on the interaction simulation between farmer behavior and policy environment under the background of human-induced disasters. And the FLPI model uses survey data as input, which could provide behavior rules closer to real situations than that of simplifying assumptions. The research indicates that the “bottom-up” approach of the FLPI model proven more appropriate to presenting micro-level adaptation and dynamic changes than traditional statistical models.And the agent-based model is very suitable for simulating the behavioral decision-making changes of farmers in policy environments.