Calibration of agricultural risk programming models using positive mathematical programming

Mathematical programming models of farmers’ cropping decisions must first be calibrated before they can be used to examine agricultural producer responses to policy changes. In this paper, we compare three calibration approaches for disentangling the risk parameter from the parameters of the cost function: one assumes a logarithmic utility function, while the others employ an exponential utility function. Historical crop insurance data for southern Alberta, Canada, are used to assess the calibration performance of the three approaches, and sensitivity analysis is implemented to test whether the changes in the optimal land allocation caused by the changes in the values of the parameters are practically reasonable. Only one of the three models is of practical use for policy analysis because it can recover the true values of the parameters and the results of sensitivity analysis are reasonable.

农户种植决策数学规划模型在用于探究农业生产者对政策变动的响应前，必须先进行校准。本文针对从成本函数参数中分离风险参数的三类校准方法展开对比：其中一类采用对数效用函数假设，另外两类则使用指数效用函数。研究采用加拿大阿尔伯塔省南部的历史作物保险数据，对三类方法的校准性能进行评估，并通过敏感性分析检验参数取值变动所引发的最优土地配置变化是否具备实际合理性。三类模型中仅有一种可应用于政策分析，因其能够还原参数的真实取值，且敏感性分析结果符合实际逻辑。