Demand System Estimation with Upward‐Sloping Supply

Abstract

This paper motivates the use of three‐stage least squares in demand systems estimation in which disaggregated data are used and supply adjustment costs exist. A system of Japanese meat expenditures is used as an example. It is shown that all but one of the meat prices are endogenous to the system. Correcting these prices for endogeneity has a significant effect on the calculated elasticities.Cet ouvrage utilise la technique de minimisation des moindres carrés à trois étapes dans un système de demande utilisant des données désagrégées et où des coûts d'ajustement de l'offre existent. Un système de dépenses pour les viandes au Japon est utilisé comme exemple. Il est démontré que tous les prix des viandes sont endogènes à l'exception d'un seul. Les élasticités qui sont dérivées sont très sensibles à l'endogénization des prix.SUMMARYThere are three important implications of this work. First, demand systems estimates that overlook supply response are as subject to simultaneous equations bias as single ad hoc demand equations. Theil shows theoretically that assuming supply curves are perfectly elastic, when in fact they are not will underestimate price responsiveness in demand equations. An empirical example is presented that demonstrates that the price elasticities generally increase when upward‐sloping supplies are assumed.Second, the iterative testing procedure presented may provide direction for model building when the true structure of the system is unknown. For example, the results of the Wu‐Hausman test indicate that assuming chicken supply is perfectly elastic in a model of the Japanese livestock industry is justified. The results also indicate that the supplies of Wagyu beef, dairy beef, pork and fish are upward‐sloping and therefore should be modeled as endogenous variables in the demand system.Third, the results emphasize the sensitivity of projections of Japanese beef imports to the assumptions underlying the demand system. If perfectly elastic meat supplies are assumed for an analysis of reducing Japanese beef import liberalization, the results will likely underestimate the impacts on beef imports.In summary, the supply curves for agricultural products tend to slope upward within the time periods used for traditional policy analysis and demand system estimation, which in turn implies that prices are determined endogenously within the system. Endogenous price determination is contrary to the assumptions that underlie the theoretical foundations and many empirical applications of demand systems. We present a methodology to test for and adjust demand systems for endogeneity. The importance of this adjustment is demonstrated by using an analysis of the liberalization of the Japanese beef market.