View article

Although the traditional (unitary) model of the household is simple, it is hard to overstate its contribution to the understanding of economic and social behavior. Whether or not it is an adequate description of household choices is fundamentally an empirical question—the answer to which likely depends on the specific application. There has, however, been little empirical testing of the model.

In this chapter I provide some evidence on this question, drawing on survey data from Brazil. Whether observed consumption and investment patterns are sensitive to differences in the distribution of income between men and women is determined in the context of household demand for commodities, nutrition, and health. Such sensitivity would be prima facie evidence against the traditional model of the household, which assumes that household decisions are unaffected by shifting resources from one member to another within a household. Rejection of the traditional model has obvious implications for public policy. It suggests, for example, mat policies that result in more resources in the hands of women will have different effects on household choices than policies that generate income for men. Previous chapters have described the theoretical models underlying the analyses presented here. In this chapter the empirical strategy adopted is outlined first, followed by a description of the data. Empirical results are presented in three sections. We examine shares of the household budget spent on a series of commodities first and then consider household nutrient demands. These outcomes are measured at the household level. The third section focuses on indicators …