Health, welfare and household resources: An analysis of Chile, the United States and Mexico

This dissertation includes three chapters, of which the first two analyze the intra-household allocation of resources, while the third one analyzes the distribution of resources across households. Using Chilean data, Chapter One presents evidence of how the nutritional status of the child reflects differences in parental preferences and child rearing technology. The results show gender specialization on child rearing in that mothers give more resources to their daughters and fathers to their sons. The education of the parents is found to be a key determinant in the assignment of resources within the family.;Chapter Two tests the unitary model of household behavior. We take the view that Aid to Families with Dependent Children (AFDC) provides a potentially powerful tool to test this model within the context of a natural experiment. If AFDC provides a natural fall-back for a low-income married women, as AFDC payments increase, separation will become more attractive and the bargaining power of the woman should increase. Using the Panel Study of Income Dynamics (PSID) for the period of 1968 to 1992, we find that households with young children, in particular those at the lower tail of the income distribution, allocate less of their income on food as maximum AFDC increases. Time allocation patterns are consistent with this finding as well. We conclude that options outside the marriage, as indicated by AFDC benefits, affect the bargaining power of women within the marriage, suggesting more resources being allocated to investment in children.;Chapter Three studies the allocation of resources across households and analyzes the change in living standards of Mexican households during the peak of the economy's transition from 1984 to 1994. We focus on changes in the standard of living using a series of income-expenditure surveys. Instead of applying conventional measures of welfare, we use graphical analysis of stochastic dominance to avoid relying on poverty lines. We include measures of sampling variation and test the significance of changes in welfare using simultaneous non-parametric statistical inference. Results show that households, increase their welfare, on average. Those with higher educational attainment of the head are more sensitive to the economy's business cycle.