Household income pooling and the demand for food: does family financial structure matter?

Research on food consumption and expenditures usually employs the use of unitarymodels that do not account for type of family financial structure. This research presentstwo collective models of household behavior, conditional and unconditional models,which were used in the analysis of the household data that came from the “ParentalTime, Role Strain and Children’s Fat Intake and Obesity Related Outcomes”. Thisresearch utilized the Generalized Method of Moments in the estimation of the system ofexpenditures on food at and away from home to test the validity of the unitary modelby testing one of the implications of this model, the income pooling hypothesis, as wellas family financial structure. It was found that the omission of family financialstructure and not the income pooling hypothesis would lead to the incorrect assertionthat the unitary model is the correct model for the analysis of intrahousehold allocation.The collective models proposed in this research were found to be preferred to those ofthe unitary models. These two models, conditional and unconditional, not only allowfor the effect of earned and unearned incomes of fathers and mothers to be different,but also incorporate family financial structure into the analysis of expenditures on food at and away from home. This research shows that the parameters of the unitary modelsare reduced form parameters that do not represent the effect that the variable of interesthas on the household expenditures category of interest. This research finds that thesereduced form parameters show the total effect which is composed of three parts. First,the change in the expenditure category of interest that comes about from a change inthe variable of interest when we hold family financial structure constant. Second, thechange in the expenditure category of interest that comes about from a change in thefamily financial structure. Third, the change in family financial structure that comesabout from a change in the variable of interest.