Abstract:
This paper presents a new model of the household that is able to explain a variety of consumption patterns that existing models cannot describe, most notably, those associated with the Deaton and Paxson (1998) paradox. The most distinctive feature of this model is the presence of common-pool goods (rival and non-excludable) previously ignored in the literature. Under regularity conditions, the model can be interpreted as a hybrid between non-cooperative and a collective models of the household. Empirically, the paper revisits the Deaton-Paxson paradox exploiting household splits in longitudinal data and computes the elusive indifference scales coefficients.