Consumer Demand and the True Cost-of-Living Index

Since Stone (1954) proposed the linear expenditure system, empirical studies on consumer demand have taken two directions. One was the effort to apply more flexible functional forms and the other was the extension of the linear expenditure system in the dynamic form. Studies by Theil (1965), Barten (1967), Christensen, Jorgenson, and Lau (1975), and Deaton and Muellbauer (1980a) belong to the first category while Stone and Rowe (1958), Pollak and Wales (1969), and Phlips (1972) belong to the latter. The flexible-form approaches have provided us with a less restrictive set of demand functions so that homogeneity an.d symmetry could be tested explicitly instead of having them as part of the maintained hypotheses. However, the rejection of these hypotheses and the violation of the negativity condition, as have been often the case in studies such as Christensen, Jorgenson, and Lau (1975) and Deaton and Muellbauer (1980a) make it difficult to go beyond testing the theory of demand. For this reason, applied welfare analysis, including areas such as the calculation of cost-of-living index and optimal taxation, has used Stone's system predominantly, despite its restrictive features. This paper follows the latter approach and applies both static and dynamic versions of the linear expenditure system to Korean data (1953-84)