The environmental consequences of agricultural production have reemerged as an important policy issue in several areas. In Florida, the US Environmental Protection Agency (EPA) is considering numeric limits on nutrient levels for Florida agriculture. The problem of missing values is becoming increasingly important in applied analysis for a variety of reasons. First, survey instruments such as the US Department of Agriculture's Agricultural Resource Management Survey (ARMS) is increasingly the only viable dataset for analyzing farm-level decisions. Estimating the response function for corn highlights several salient difficulties. Aggregate sampling implies that when one input is missing, all input variables are likely to be missing. The reduced quality of fit in the multiple imputations may involve the relative value of information on exclusion. The pseudodata results indicate that appropriate estimates of the production function can be generated from the application of these procedures.
ASJC Scopus subject areas
- Agricultural and Biological Sciences (miscellaneous)
- Economics and Econometrics