ALTERNATIVE ESTIMATION METHODS OF NONLINEAR DEMAND SYSTEMS

Several contemporary models of consumer demand comprise complete sets on nonlinear demand functions. Estimation methods should take into account parameter nonlinearity, cross-equation correlation, variance-covariance singularity of the disturbance terms, and various parameter restrictions. This paper presents a theoretical discussion and some empirical results using the maximum likelihood (ML) method and the iterative version of Zellner's seemingly unrelated regression (IZEF) method in the estimation of a nonlinear system of demand equations (the linear expenditure system) when the disturbance terms are both contemporaneously and serially correlated. On the basis of the evaluation of parameter estimates and their asymptotic standard errors as well as the cost of computation effort, the IZEF technique is preferred over the ML technique in this empirical problem.