The Intrahousehold Demand for Nutrients in Rural South India: Individual Estimates, Fixed Effects, and Permanent Income

Good estimates of nutrient intake responses to prices and income are very useful for the evaluation of the numerous efforts to improve nutrition in many developing countries through price-subsidy and income-generation policies. We discuss three problems in standard estimates of these responses and then illustrate their implications for nutrient demand relations for a poor sample from rural south India. (1) Intra-household nutrient allocations usually are ignored. In this case nutrient intakes for females systematically have algebraically lower price elasticities than do those for males, which may leave the females particularly vulnerable at times of food shortages. (2) Unobserved fixed effects may bias the estimates of responses to observed variables. In this case not only the community fixed effects on which the previous literature has focused, but also household and individual fixed effects are important. Failure to control for them results in substantial algebraically upward biases in many estimated price responses. (3) Most previous studies use current instead of permanent income, which a priori may account for the low estimated income elasticities. In this case, however, the use of permanent income does not change the conclusion that the nutrient intakes responses to income are quite small.