Given the importance of food products in households' basket, the study carries out a systematical analysis of the urban households' food expenditure during 1983-2000. In this system (AIDS), the food cost share has been considered as a function of the given foodstuff price, the price of all other foodstuff and allocated real income. To design the model, the food stuff has been divided into 8 categories: 1. rice, flour, its processed products, 2. meat (beef), 3. meat (chicken and fisheries), 4. dairy products, 5. vitamins, 6. animal oil, 7. herbal oil and 8. all other food products.
Using economic variables, based on microeconomic theory, the cost share and the cost and income elasticity are estimated The parameters are estimated by applying the pooled time series and cross-section data and seemingly unrelated regression techniques and the stationary of the model variables are examined through the Augmented Dickey Fuller (ADF) test. Taking into view the homogeneity and symmetry test the authors go on to specify the model by the Error-Correction Method. As the households' decisions about the food expenditure are made in a two-stage process, the two-stage budgeting technique is employed.
The findings show that the income elasticity for food is lower than unity and for other goods is greater. Also, in foodstuff category, the own-price elasticities have correct sign and the income elasticities for rice, flour, and its products, dairy products, herbal oil and vitamins are lower than unity and for the other foodstuff are greater.