Time series analysis of Generalized Cobb-Douglas production technologies with serial, correlated stochastic technical change : an application to regional agricultural production

Abstract

The focus of this research is on empirical characterization of technical relationships among aggregate factors in regional agricultural production processes. The Generalized Cobb-Douglas (GCD) functional specification is employed to relax strict neoclassical efficiency constraints and provide a means for modeling systematic and stochastic technical change without strict reliance on time trend variables. Two GCD class models are specified under a first order autoregressive schema: The Constant Elasticity of Marginal Rate of Technical Substitution model (GCD-CEMRTS) and the Constant Ratio of Elasticity of Substitution model (GCD-CRES). Each model is tested against commonly used classical functional forms which have technology images within their respective GCD class. Likelihood ratio tests indicate strong support for rejection of the classical forms relative to their unrestricted GCD class counterparts. In conclusion, gross factor demand elasticities are constructed and analyzed to determine technological relationships among factors and specified systematic technology changers.