Estimation of consumer demand function in low and high-income classes in Iran urban areas: An application of pooled data approach in AID system

This article aimes to estimate consumer demand function using Dynamic Linear Almost Ideal Demand System and Seemingly Unrelated Regression (SUR) method in 1982-2007 period. Seven income classes of urban areas in Iran are selected as cross-sections and then using the estimated own-price and cross-price elasticities, we try to compare the consumption behavior of households in the several income classes under consideration. To undertake this task, several hypotheses includeding homogeneity and symmetry hypothese are tested. Results show that the consumers in low-income class face with money illusion, meaning that they are insensitive to their real income. Again, high-income class consumers are not sensible to their real incomes when food and miscellaneous goods are under investigation, while this is not the case when they consume clothing and housing goods. Results also confirm the assymmetry hypothesis in the consumer behavior. The examination of demand-law hypothesis show negative own- price elasticities in all commodity groups, as is expected. Results of estimation of income elasticities indicate that food and housing groups are essential but miscellaneous and clothing groups are luxury, when the low-income group is under consideration. For high-income group, hygienic commodities, found to be more essential than all of the other commodity groups are. Moreover, survey of cross-price elasticities show that for the same commodity groups, substitute and supplementary properties are different in various income groups.