Simulating Heterogeneous Farmer Behaviors
formation “ nudges ” in NPS pollution management . We are interested in if information “ nudges ” based on social comparison and peer action could help the performance of ambient-based policies . In the first information treatment , participants are provided with information about what people “ like them ” have chosen in a similar situation in the past . In the second treatment , their group in the preceding round provides participants with information regarding average production and average rate of adoption of technology . Participants ’ responses to the policy and the information treatments given the heterogeneity of production types are used to guide the agent ’ s behavior in the models under various scenarios .
In this study , we scale up findings from an economic experiment with ABM in a spatially explicit watershed setting to provide insight into the effects of different policy interventions addressing NPS pollution . The models capture interactions among heterogeneous agents in terms of diffusion of technology adoption by farmers , which is difficult to model using other techniques . Specifically , we test how tax / subsidy policies based on ambient levels of water pollution work in scenarios involving heterogeneous production and pollution schemes and focus on cases in which the decision space of the agents is extended from making a single production decision to making a production and a technology decision . We also investigate how information influences people ’ s behavior and whether policies can be designed to incorporate information “ nudges ” to induce more-desired outcomes . Our study contributes to the literature in two main aspects . In environmental and resource economics , our experiment investigates the effect of information “ nudges ” in an experimental setting that simultaneously incorporates an extended participant decision space and multiple layers of heterogeneity . Moreover , we use an ABM that features heterogeneous agents in a spatially explicit context to understand implications of the complex actions and interactions created based on experimental data . In the field of ABM , despite rising interest in using non-fully rational agents , not much work has actually done so . We are one of the first to introduce bounded rational agents into an ABM based on an economic experiment . The ABM agent decision rules are closely linked with human decisions in the economic experiment using an underlying game-theoretical model . Our research demonstrates that economic experiments can be useful to capture bounded rationality and guide ABM development . This study provides an example to incorporate human-based decision rules and a possible framework to integrate experiments and ABM in future research .
Experimental Design and Theoretical Foundation
In this part , we discuss the experimental design of our economic experiment . We first lay out the theoretical model , and describe the treatments in the experiment .
We build upon and extend the classic model framework in the environmental economics literature . Consider a group of agricultural producers indexed by i = 1 … N operate farms or ranches adjunct to a common watershed . The farmers ’ operations generate pollution as byproduct . A regulator monitors water quality by a sensor at the downstream of the watershed . The farms may differ in both their capacity and their distance to the sensor . The farmers may choose to adopt a pollution abatement technology ( e . g ., buffer , cover crop ) at a cost ( τ ) proportional to farm size . Each year , the farmers make two decisions :
168