Food demand and the nutrient intake of households in underdeveloped rural regions of China : an instrumental variable approach

Purpose The purpose of this paper is to evaluate the effects of price and income changes on food and nutrient demand of rural households by including own-produced food and production-side effects in the demand estimation to correct potential measurement bias in the income and price elasticities for rural households in underdeveloped areas. Simulation results of income and grain price changes on food and nutrition security are provided for economic nutrition security policy applications. Design/methodology/approach This study analyzes survey data of 1,555 households from underdeveloped rural areas of China to find out how price and income changes affect food and nutrition insecurity of rural households. The authors employ the quadratic almost ideal demand system (QUAIDS) in a two-stage budgeting framework, using quality adjusted prices that were retrieved with regressions of the difference between the unit value surveyed at household level and its village average on household characteristics. The bias correction is implemented by using an augmented IV (instrumental variable) method, in which each market price is instrumented with farm-specific variables. Important macro- and micronutrient elasticities are computed for (a) households with agriculture as main income and (b) other households (of which still many have agriculture as a side business). Finally, the authors use these elasticities to simulate how changes in income or grain prices affect the food and nutrition security in the studied areas. Findings In general, food income elasticities of agricultural households are at a higher level than those for other households, and so are the food price elasticities. Income changes also have a greater nutritional effect on agricultural households than on other households. Nutrient income elasticities ranged from 0.22 (energy) to 0.27 (Vitamin A) for agricultural households and from 0.19 (energy) to 0.23 (Vitamin A) for other households. Grain price increases have greater effect on nutritional status of non-agricultural households, while a grain price reduction is not clearly favoring the nutritional situation of a particular household group. Originality/value This demand study contributes to the literature by taking into account differences in consumption of own production between households and the potential endogeneity of prices resulting thereof. The authors also demonstrate that merely reporting nutrient elasticities might not be sufficient for policy recommendations, and simulations should be reported as a valuable addition.