Sampling Schemes for Policy Analyses Using Computer Simulation Experiments

Evaluating the environmental and economic impacts of agricultural policies is not a simple task. A systematic approach to evaluation would include the effect of policy-dependent factors (sue as tillage practices, crop rotations, and chemical use) as well as the effect of policy-independent covariates (such as weather, topography, and soil attributes) on response variables (such as amount of soil eroded or chemical leached into the groundwater). For comparison purposes, the effects of these input combinations on the response variable would have to be assessed under competing policy scenarios. Because the number of input combinations is high in most problems, and because policies to be evaluated are often not in use at the time of the study, practitioners have resorted to simulation experiments to generate data. But generating data from simulation models is often costly and time consuming; thus, the number of input combinations in a study may be limiting even in simulation experiments. In this paper, we discuss the problem of designing computer simulation experiments that require generating data for just a fraction of the possible input combinations. We propose an approach that is based on subsampling the 1992 National Resources Inventory (NRI) points. We illustrate the procedure by assessing soil erosion in a situation where there are "observed" data (reported by the Natural Resources Conservation Service (NRCS)) for comparison. Estimates for soil erosion obtained using the procedure we propose are in good agreement with NRCS reported values.