Household Food Expenditures, Parental Time Allocation, and Child Overweight: An Integrated Two-Stage Collective Model with an Empirical Application and Test

Presenter: Wen You, Virginia Tech

Abstract

Rationale: The increased prevalence of childhood obesity is a major concern for the society. The factors affecting the childhood obesity are many and not well understood because children's choice sets are highly influenced by parental decisions. In the standard nutrition literature, obesity is a function of the balance between energy intake and energy expenditure (Hoffman and Sawaya). Those two components of the balance equation are influenced by genetic factors and environmental factors. No single discipline can achieve thorough investigation on this complicated issue.

Objectives: The objective of this paper is to develop an inter-disciplinary theoretical model that will provide theoretical guidance to the empirical research on children's obesity-related health outcomes.

Methodology: This paper develops a two-stage collective household production model for household food expenditures, parental time allocation, and childhood overweight. The model fills four gaps in the literature: 1) the “blackbox” treatment of intrahousehold decision-making; 2) excluding the child’s decision input; 3) excluding non-economic variables; 4) no theory-supported instrument identification. The traditional unitary household production model with children is a special case of the model developed. The theoretical framework provides a general triangular system of equations for empirical analysis.

Results: Using a unique dataset that includes all the covariates needed for the estimation, the traditional unitary model is rejected relative to the collective model developed. However, the major practical differences between the unitary and collective model are located in the parental input demand equations, not in the child overweight equation. Consequently, the unitary model would be very misleading for trying to draw inferences on input demands, such as food expenditures and time allocations, but would be generally consistent with a collective model for drawing inferences on a child’s overweight. Furthermore, in terms of the derived reduced form effects, the unitary model in general identifies the most important types of variables and provides results that are consistent with the more data intensive collective model, with a few exceptions. The theoretical and empirical framework developed in this paper should prove attractive for future work looking to model and understand the relationship between food consumption, time allocation, and health outcomes, as these are obviously interrelated activities that should be modeled together if possible. Of course, as with most new theoretical and empirical frameworks, new questions are generated as byproducts that require further work.