Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Economics / Economic Policy


M.J. Osborne




The first two Essays in this thesis deal with the issue of intra-household allocation of resources among children in India. In the first Essay, we present an analysis of gender patterns in consumption within the household, utilizing a unique panel of data on rural households. We find that female children get fewer resources than male children, as reported in the literature, by estimating Engel curves for various goods. We then hypothesize that this may be due to a wealth-effect associated with female children, since in India parents spend a considerable amount of resources on the marriages of their daughters. One more female child in the household increases the marginal utility of wealth in this context, since female children reduce life-time wealth. Invoking the assumption used in modern theories of life-cycle consumption that households seek to equalize the marginal utility of wealth over time, and utilizing the panel nature of our data, we model the (marginal utility of) wealth-effect as an unobserved household-specific fixed effect that stays constant over time. We then estimate our Engel curves with controls for the unobserved fixed effect to find that the bias in favour of male children that we observe in our levels equations vanishes in the fixed-effects version. This result supports our prediction that the bias against female children that we observe in the data is due mainly to the wealth-effect (or high marriage costs of female children). In Essay 2, we formalize the ideas presented in Essay 1 with a simple three-period model of intra-household allocation. We allow for the possibility that parents may have perferences towards male children, as an alternative to the wealth-effect motive outlined above. The model yields many intuitive and testable predictions on changes in household consumption around the birth of children. Based on the same data used in the first Essay, we find that for the wealthy households in our sample the wealth-effect motive dominates parental preferences. These households reduce their total expenditures soon after a female birth, in order to save for the anticipated marriage costs in the future. our analysis of the data from a marriage survey for the same households shows that the savings estimates that our results suggest match up with the marriage costs for female children in these wealthy households. Our results for the poorer households in the sample weakly support the parental preferences hypothesis. The last, self-contained, Essay presents a theory of urban-rural remittances in developing countries. The framework here is similar to one in the literature on the economics of the family: the interaction between the rural family and the (self-interested) migrant is formalized as a simple game. We show that migrants remit only if remittances are used in augmenting the family bequests. it is also shown that they remit only if their incomes fall within a certain interval, given the rural incomes: intuitively, rich and poor migrants do not remit because their behaviour is conditioned by parental bequests. Our comparative static predictions show that remittances are non-monotonically related to the migrant's and rural incomes, and the family's initial stock of bequests.

McMaster University Library

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Economics Commons