Generalized Cross-Entropy Estimation of a Varying-coefficients Model of Cost Allocation in Multi-Product Farming

Abstract

Standard farm-accounting information is typically restricted to aggregate input expenditures, without revealing the allocation of these expenditures to the various farm enterprises. Given the diversified enterprise base of most farm businesses, a proper assessment of each enterprise's contribution to overall farm performance is severely obscured. This paper presents a new approach for estimating the unobserved input-output or (preferably) cost-allocation coefficients for multi-product farms. Specifically, a model with (non-randomly) varying coefficients is developed, which also allows for a more realistic appraisal of the "real-world" diversity of farm operations. The coefficients of the model are estimated using the Generalized Cross-Entropy (GCE) method. The GCE method proves to be an effective way to obtain unique estimates of farm-specific cost-allocation coefficients and to overcome the many practical problems usually encountered in earlier empirical works. The proposed GCE estimator is applied to a set of 2000-2001 accounting data for a small sample of beef-dairy farms located in Brittany, France.