Accounting for structural changes in demand for foods in the presence of age and cohort effects: the case of fresh fish in Japan

Several studies indicate that not only age but also cohort effects exist in consumption of selected food products, although the latter are rarely explicitly specified in the ordinary econometric analyses of food consumption. These two demographic variables are significant in magnitude and often differ in pattern. Where the population is aging rapidly, as in Japan and Korea, it is inevitable that “structural changes” take place in the demand for certain products, because of the change in the composition of consumers. In this article, we first derive individual consumption of fresh fish, traditionally a major source of animal protein in Japan, by age from the Family Income and Expenditure Survey, where the original data are classified by age groups of household head. A cohort table based on the individual consumption estimates is then decomposed into age, cohort and period effects first and then the period effects are regressed against prices and income to determine the economic demand elasticities. We also attempted an alternative approach: a cohort analysis augmented with economic variables, to determine economic demand elasticities in one step. Even after age, cohort, price, and income effects are accounted for, non-negligible residuals remain unexplained, which may need interdisciplinary investigations.