This paper presents a structural model of intrahousehold allocation, analyzing how income and other factors influence household expenditure decisions. The authors argue that treating households as single decision-makers is flawed, as household members' relative incomes and ages significantly affect final allocation decisions. Using Canadian data on couples without children, they find that expenditures on each partner depend on their relative incomes, ages, and household wealth. The paper critiques the unitary model, which assumes households behave as single agents, and shows that empirical evidence contradicts this assumption. Instead, the authors propose a "collective" model where household decisions result from interactions between individuals with different preferences. They develop a parametric model of intrahousehold allocation, showing that conventional family expenditure data can be used to identify allocation decisions under certain assumptions. The authors test the unitary model against empirical data and find that it fails to explain household behavior, particularly for couples. They develop a model where each person's consumption is treated as an exclusive good, and show that the allocation of these goods depends on the relative incomes and ages of the partners, as well as household wealth. They estimate the parameters of this model using Canadian data on couples with no children, finding that "who gets what" in the household depends on these factors. The paper also discusses the theoretical framework for modeling intrahousehold allocation, including the distinction between private, public, and exclusive goods, and the mechanisms used to reach decisions. It presents a sharing rule model, where total household income is allocated to saving, public goods, and each partner's expenditure on nonpublic goods. The authors show that this model can be identified using conventional family expenditure data, and that the sharing rule reflects the outcome of the decision process. The paper concludes that the unitary model is insufficient to explain household behavior, and that a more realistic model must account for the interactions between individuals with different preferences. The authors argue that the collective model is more appropriate for analyzing intrahousehold allocation, and that the findings support the use of structural models to understand how income and other factors influence household decisions.This paper presents a structural model of intrahousehold allocation, analyzing how income and other factors influence household expenditure decisions. The authors argue that treating households as single decision-makers is flawed, as household members' relative incomes and ages significantly affect final allocation decisions. Using Canadian data on couples without children, they find that expenditures on each partner depend on their relative incomes, ages, and household wealth. The paper critiques the unitary model, which assumes households behave as single agents, and shows that empirical evidence contradicts this assumption. Instead, the authors propose a "collective" model where household decisions result from interactions between individuals with different preferences. They develop a parametric model of intrahousehold allocation, showing that conventional family expenditure data can be used to identify allocation decisions under certain assumptions. The authors test the unitary model against empirical data and find that it fails to explain household behavior, particularly for couples. They develop a model where each person's consumption is treated as an exclusive good, and show that the allocation of these goods depends on the relative incomes and ages of the partners, as well as household wealth. They estimate the parameters of this model using Canadian data on couples with no children, finding that "who gets what" in the household depends on these factors. The paper also discusses the theoretical framework for modeling intrahousehold allocation, including the distinction between private, public, and exclusive goods, and the mechanisms used to reach decisions. It presents a sharing rule model, where total household income is allocated to saving, public goods, and each partner's expenditure on nonpublic goods. The authors show that this model can be identified using conventional family expenditure data, and that the sharing rule reflects the outcome of the decision process. The paper concludes that the unitary model is insufficient to explain household behavior, and that a more realistic model must account for the interactions between individuals with different preferences. The authors argue that the collective model is more appropriate for analyzing intrahousehold allocation, and that the findings support the use of structural models to understand how income and other factors influence household decisions.