Evolving food consumption patterns of rural and urban households in developing countries: A Bangladesh case

Published: 10-03-2019| Version 1 | DOI: 10.17632/zvmt38vy3n.1
Khondoker Mottaleb


This dataset is consists of Bangladesh’s Household Income and Expenditure Survey (HIES) data collected in 2000, 2005 and 2010 by the Bangladesh Bureau of Statistics (BBS). In the first stage, the BBS selects primary sampling units (PSUs) consisting of specific geographical locations, and in the second stage, it randomly selects 20 households from each PSU that represent rural and urban areas. In HIES 2000, BBS selected 442 primary sampling units and surveyed 7,440 households of which 5,063 were from the rural areas and the rest were from urban areas. In HIES 2005, BBS selected 504 primary sampling units and surveyed 10,080 households of which 6,040 were from the rural areas, and the rest were from urban areas. In HIES 2010, 1,000 PSUs were selected and they surveyed 12,240 households of which 7,840 were from the rural areas and the rest were from urban areas. The HIES 2000, 2005 and 2010 data on household-level consumption is quite detailed. The consumption of food items both in quantity and expenditure was divided into 17 major food categories. The households were asked to describe their consumption habits during a two-week period. The major food categories were cereals, pulses, fish, eggs, meat, vegetables, milk and dairy products, sweets, oil and fats, fruits, drinks, sugar and molasses, tobacco and related items, spices and betel leaves, including betel nuts. An analysis of a complete demand system consisting of separate equations for all items in 17 major food categories would require a substantial amount of computer memory and time. For simplicity, the present study analyzed the demand systems considering a subset of food items commonly consumed by the sampled households: cereals, pulses, fish and vegetables. Among the cereals, rice (grain), and wheat (flour) and other rice and wheat products, such as rice cakes, biscuits and miscellaneous rice and wheat products are separately treated. As the five food categories are the most commonly-consumed items, and as the others are sporadically-consumed food items, policy implications based on the present study can provide deep insights into the evolution of the basic food consumption pattern in Bangladesh by the rural-urban affiliation of the households. Note that in HIES 2000 the expenditure information on vegetables for 84 households was missing; in HIES 2005 information on vegetable expenditure for 783 households was missing, and such information was missing for 1,229 households in HIES 2010. Thus, the dataset includes 28,384 households from HIES 2000, 2005 and 2010, of which 19,471 are rural households and 8,913 are urban households.


Steps to reproduce

Requires Stata 15.0 or above level version. Just follow the steps described in the do.file.