GLOBALLY CONVEX AGRICULTURAL PRODUCTION SYSTEM - PARAMETER-ESTIMATION

The goal of incorporating, into a computable general equilibrium model , an econometric model of agricultural production that maintains flexi bility and global convexity for a large number of inputs and outputs l eads to a computationally complex estimation problem. We have develope d and implemented an iterative least squares procedure to estimate the parameters of this nonlinear production system subject to nonlinear c onstraints imposed by the underlying economic theory. The large comput er memory of the Cornell National Supercomputer Facility allows the es timation of this large, globally convex, flexible agricultural product ion system.