Testing for Behavioral Objective and Aggregation Opportunities in U.S. Agricultural Data

To guide formulation of an aggregate model of U.S. agricultural production, this paper reports nonparametric test results for both behavioral objective and aggregation level. Nonparametric tests are implemented as heuristics to determine empirical support for simplified analytical structures. Evidence does not refute the joint hypothesis that producers behave collectively as though they are a price-taking, profit-maximizing firm with a convex technology subject to nonregressive technical change. The empirical results also support exhaustive aggregation of input categories, but much less aggregation of outputs. This paper reports the first known implementation of the nonparametric homothetic separability tests. Copyright 1998, Oxford University Press.