Estimating price and income elasticities in the presence of age-cohort effects

Individual consumption of most food products varies by age, and in countries like Japan that have experienced drastic social and economic changes during the past several decades, it also differs from generation to generation. Unless proper measures are taken to account for these factors, estimates of demand elasticities could be severely biased. In this study, individual consumption of fresh fruit was derived from Japanese household data classified by age of household head for the years 1979-2001. Individual consumption was then decomposed by age, cohort, and period effects using Bayesian cohort analysis. Pure period effects thus determined were regressed against changes in price and income, to obtain less biased estimates for demand parameters than non- or partially age-compensated analysis. [Econlit citations: Q110]. © 2006 Wiley Periodicals, Inc. Agribusiness 22: 201-217, 2006.