Abstract:

I study distribution of resources within households via a structural model of household demand, using 14 years of the UK Family Expenditure Survey. I focus specifically on allocation to children versus adults. My study differs from earlier analyses in two important respects.

First, I specify a complete demand system for effectively all household expenditures, which allows me to study not only the effect of marginal increases in income, but also fully- specified substitution effects across goods. I show that there is substantial bias if one estimates a demand function only over a subset of goods, as is common.

Second I develop a GMM-based technique for estimating the demand functions that does not require any assumptions on the orthogonality of income and preferences. The main result is that a disproportionate share of the marginal dollar of income is expended on adult-assignable and leisure goods relative to the average dollar. Adult and leisure goods are last-doll ar goods, whereas children’s and joint household goods are first-dollar goods. Further, relative price decreases of joint household or child-assignable goods induce substantial increases in purchases of adult-assignable and leisure goods. Conversely, price increases of adult-assignable goods lower consumption of child-assignable goods.

Back to Schedule