Estimating Adoption of Cover Crops Using Preferences Revealed by a Dynamic Crop Choice Model

In this paper, we propose a framework based on micro-level dynamic land use models to predict the adoption of cover crops in the Upper-Mississippi River Basin. We use preferences recovered using a dynamic discrete model of crop choice to build a dynamic optimization framework to evaluate a range of scenarios based on the cover crops’ effect on cash crop yields, costs of cover crop operations, and government support. We use a conditional choice probability method to estimate the dynamic crop choice model based on field-scale panel data and value function iteration method to assess counterfactual cover crop scenarios. The framework is expected to be applicable to modeling decisions to adopt conservation technologies in the absence of individual-level adoption data or for cases when conservation technology is new. The dynamic crop choice model yields expected results and reveals preferences for net revenues in line with previous literature. Simulation results predict baseline cover crop adoption rates which, although in line with some recent farmer surveys, are quite a bit higher than rates reported in 2012 Census of Agriculture. We attribute these results to substitution patterns implied by a dynamic logit model estimated, and suggest using aggregated Census of Agriculture data on state-wide adoption of cover crops to calibrate the constants in the the estimated dynamic logit model as a possible remedy under paucity of data related to individual decisions on cover crops adoption